Illegal related-party transactions (RPT) are federal felonies that pose a severe threat to the stability and integrity of modern financial systems. The increasing frequency of RPTs forms complex and dynamic networks. Existing temporal graph learning methods tend to treat entities as functionally homogeneous, ignoring the diverse and evolving structural roles of nodes. Role-based embedding methods model global structure by bridging same-role nodes, but their reliance on a unified mechanism for aggregation and evolution means they fail to distinguish the underlying logic of distinct interactions governed by structural roles. The limitations motivate us to develop a customized role-based strategy. It can also adapt to evolving RPT dynamics, thereby forming a continuous regulatory process to combat illegal activities. In this paper, we propose an innovative Role Perceptual Augmented Temporal Graph Network (RPATGN) for proactive RPT detection. We analyze the structural roles of nodes and employ a role-based message passing mechanism that adapts its aggregation strategy based on the roles of interacting nodes. We integrate a variational graph recurrent neural network, enhanced by temporal contextual attention, to explicitly model the dynamics of the roles and the overall network evolution. Extensive experiments on real-world financial datasets demonstrate the effectiveness of our approach for RPT detection. It holds practical significance for fostering robust financial systems and promoting healthy, transparent financial markets.

Graph fraud detection (GFD) on transaction networks is crucial for safeguarding financial systems. However, due to the limited perspective of existing graph neural networks (GNNs) in the single transaction view, sophisticated fraudsters can disguise themselves to exhibit weak fraud signals, appearing as borderline fraudsters. To address this challenge, we propose MH-LGC, a multi-view hypergraph fraud detection model with large language model (LLM) guided contrastive learning. MH-LGC tackles two key limitations of existing GNN-based GFD methods: (1) Due to the local aggregation mechanism, existing methods struggle to capture high-order trading patterns among distant fraudsters. MH-LGC introduces two temporal hyper-views as complements to the transaction view and employs a Temporal Hypergraph Attention Network (THAN) to integrate the three views. (2) Most GFD methods overlook the rich semantic cues embedded in transaction data. Although some general graph learning studies have explored LLM integration, the high computational overhead and task-specific fine-tuning make them impractical for GFD tasks. MH-LGC introduces a semantic view through a fine-tuning-free LLM-Guided Contrastive learning (LGC), adopting a novel paradigm for integrating GNN and LLM to reduce the computational overhead of LLM. Extensive experiments on three real-world datasets demonstrate that MH-LGC outperforms twelve state-of-the-art baselines, with AUC improvements ranging from 1.10% to 5.70%.

Fraud detection on streaming transaction graphs (STGs) faces challenges on the catastrophic forgetting of previously learned fraud patterns when adapting to evolving patterns. Although some Graph Continual Learning (GCL) approaches mitigate this issue by storing and revisiting historical samples, practical storage constraints prevent them from fully preserving previous patterns. In this work, we propose STG-DGR, a streaming GNN model with diffusion-based generative replay that generates synthetic samples to retain previously learned patterns without storing real samples. The generation of replay samples for STGs faces two key challenges: (1) Heterogeneity challenge of generating STG samples with discrete adjacency table, user features, transaction features, and transaction timestamps. (2) Dependency challenge of capturing bottom-up dependencies across layers in STG samples. To address these challenges, STG-DGR integrates two novel components: (1) a Computational Subgraph Processor (CSP) that transforms heterogeneous STG samples into well-organized hierarchical subgraphs, and (2) a Diffusion-based Subgraph Generator (DSG) that captures the bottom-up dependencies using a novel Transformer-based Hierarchical Denoising Network (THDN), and generates synthetic replay samples that preserve these dependencies. Extensive experiments on four streaming fraud detection datasets demonstrate STG-DGR's superiority in reducing forgetting and improving accuracy over nineteen state-of-the-art baselines.

In the rapidly evolving web-based financial ecosystem where digital banking services become critical infrastructure for underserved communities, credit card fraud disproportionately affects vulnerable populations relying on financial platforms. However, previous studies overlook extreme data scarcity conditions, particularly at small-to-medium web banks that serve as crucial gateways for vulnerable communities. This paper addresses the fundamental challenge of building inclusive and secure financial systems operable at true web scale. To overcome this deficiency, we propose a novel web-based fairness-aware federated fraud detection model, CLARF, which utilizes the designed privacy-enhanced representation fusion and fraud-aware contrastive learning modules to enhance detection performance under conditions of data scarcity and label imbalance. Furthermore, current federated fraud detection systems critically neglect vulnerability to backdoor attacks, where malicious actors can implant hidden triggers during model aggregation, compromising system integrity. We propose a novel dynamic web Min-Max adversarial game framework where attackers employ hybrid multi-stage reinforcement learning with multi-dimensional reward mechanisms to dynamically evolve triggers that achieve excellent tradeoff between stealthiness and effectiveness. Defender adapts a closed-loop Selection-Evaluation-Suppression framework where high-reliability clients are selected via Fisher information to carry out reverse trigger engineering. Then clients' confidence scores are calculated as weights to minimize Attack Success Rate (ASR) during aggregation. Extensive experiments on six financial fraud datasets demonstrate the superiority of CLARF model and Min-Max adversarial game paradigm compared with multiple SOTA models.

Video Question-Answering (VideoQA) enables machines to interpret and respond to complex video content, advancing human-computer interaction. However, existing multimodal large language models (MLLMs) often provide incomplete or opaque explanations and existing benchmarks mainly focus on the correction of final answers, limiting insight into their reasoning processes and hindering both transparency and verifiability. To address this gap, we propose the Question Parsing, Video Alignment and Answer Aggregation framework (QPVA$^{3}$3), which leverages a compositional graph to drive visual and logical reasoning in VideoQA. Specifically, QPVA$^{3}$3 consists of three core components, the planner, executor, and reasoner to generate the compositional graph and conduct graph-driven reasoning. For the original question, the planner parses it into the compositional graph, capturing the underlying reasoning logic and structuring it into a series of interconnected questions. For each question in compositional graph, the executor aligns the video by selecting relevant video clips and generates answers, ensuring accurate, context-specific responses. For each question with its first-order descents, the reasoner aggregates answers by integrating reasoning logic with visual evidence, resolving conflicts to produce a coherent and accurate response. Moreover, to assess the performance of existing MLLMs in the reasoning processes of VideoQA, we introduce novel compositional consistency metrics and construct a VideoQA benchmark (QPVA$^{3}$3 Bench) with 3,492 question-video tuples, each annotated with detailed compositional graphs and fine-grained answers. We evaluate the QPVA$^{3}$3 framework on QPVA$^{3}$3 Bench and 5 other VideoQA benchmarks. Experimental results demonstrate that our framework improves both consistency and accuracy compared to baselines, leading to a more transparent and verifiable VideoQA system. This approach has the potential to advance the field, as supported by our comprehensive evaluation and benchmarking efforts.

Graph Contrastive Learning (GCL) methods typically leverage augmentation techniques to generate different graph views for comparison, thereby learning corresponding representations for graph-related tasks in label-scarce scenarios. However, existing GCL methods suffer from two primary limitations: 1) they use predefined or one-time perturbations for augmentation, ignoring adaptive noise injection during forward propagation and thus leading to suboptimal model robustness; 2) their contrast mechanisms mainly focus on the agreement of inter-graph representations while neglecting the dimensional feature redundancy within intra-graph representations. To solve these issues, we propose Layer-adaptive-augmentation-based Graph Contrastive Learning with feature Decorrelation (LGCLD). First, the designed layer-wise adaptive augmentation method performs dynamic perturbations while maintaining the semantic similarity between augmented and original graphs, which can improve model robustness. Second, we introduce an Agreement-Decorrelation loss (AD loss) that simultaneously optimizes the agreement between graph-level representations and the feature correlation among different dimensions within each graph-level representation, promoting the model to learn informative and non-redundant graph-level representations. Furthermore, we analyze the reasonableness of AD loss through the graph information bottleneck principle. Experiments on various-domain graph datasets demonstrate that LGCLD achieves better or competitive performance compared with a series of state-of-the-art baselines.

Money laundering is the process of hiding the origin of illicit funds to make them appear legitimate, thereby threatening the integrity of financial systems. To detect money laundering activities, graph neural networks (GNNs) have been widely adopted to model complex relational structures in transaction networks. However, a closer inspection of real-world money laundering cases reveals that launderers deliberately establish connections with multiple licit accounts to mask their illicit attributes. Such disguising behavior introduces network heterogeneity, which contradicts the fundamental assumption of homophily for most GNNs. Additionally, money launderers further conceal their activities by obscuring illicit fund flows through multihop transaction paths. This strategy poses a significant challenge for GNNs, as their limited receptive fields struggle to capture such long-range dependencies. To address these challenges, we propose a dual-view graph representation learning method, named DC-LCG, to detect money laundering. DC-LCG employs complementary local and contextual views to expose disguises and trace illicit flows, respectively. The local view implements a soft-label-guided dynamic grouping and aggregation method that separates nodes into illicit and licit groups, performing probability-weighted aggregations to mitigate network heterogeneity and expose disguises within transaction networks. The contextual view employs a dynamic path pruning method to filter licit nodes and enhance paths relevance, followed by multipath semantic fusion through transformer-based encoding to capture long-range dependencies across multihop transaction paths. A mutual attention mechanism integrates both views to create comprehensive node representations. Experiments on three public transaction datasets show that DC-LCG outperforms state-of-the-art baselines by 2%–10% across evaluation metrics.

Credit card fraud brings serious losses to both cardholders and card issuers. To reduce losses caused by fraudulent behaviors, banking institutions establish credit card fraud detection (CCFD) models to identify potential fraudulent behaviors. To develop more effective fraud detection models, banking institutions need to collaborate on model training. Federated learning (FL) enables collaboration on fraud detection model training without exchanging data between banking institutions. Nevertheless, the distribution of transaction data in the real-world is heterogeneous among banking institutions, which may lead to convergence issues in global fraud detection models. Furthermore, the behavior and weights of the model may implicitly contain the cardholders’ personal information, which makes existing federated models prone to transaction data leakage. In this article, we propose aheterogeneous credit cardfrauddetection model based onfederated learning withmembership privacy, called HFDFM. To protect the sensitive information of cardholders, we design a novel mechanism that ensures training-data confidentiality by minimizing the accuracy of the best black-box membership inference attack (MIA) against the model. Unlike previous FL frameworks that either optimize for client drift or for membership privacy, HFDFM simultaneously mitigates drift and provides certified membership privacy through a single min–max game. Additionally, we use the control variable to rectify the client drift in its local update in the scenario of heterogeneous data. Extensive experimental results on three mainstream real-world transaction datasets demonstrate that the proposed HFDFM has advantages in utility compared to 11 SOTA baselines, and the proposed HFDFM can mitigate the risks of MIAs (near random guess).

Nowadays, cooperation between different organizations to form a cross-industry anti-fraud alliance by using federated learning (FL) is promising for transaction fraud detection. However, there are two challenges when building the federated learning-based fraud detection method. First, existing FL-based fraud detection methods cannot fully use participants' local data to better distinguish unknown types of abnormal fraud transactions. A lot of local data would be discarded due to the private set intersection process under the federated learning framework. Second, they mainly focus on the fraud detection performance but ignore the privacy and communication costs during the training of the fraud detection model. In response to the above challenges, we propose a privacy-aware (P) transaction fraud detection model with low (L) communication costs (C) under a vertical and horizontal joint federated learning framework in this paper, named PLC. Specifically, we merge horizontal federated learning and vertical federated learning to build a normal behavior model to improve the fraud detection performance for solving the first challenge. Meanwhile, we design a sparse factor-based SMC aggregation method to reduce communication costs without sacrificing detection performance for solving the second challenge. The experimental results show that the communication costs can be reduced by 46.7% and 55.4% under the ECC dataset and the Vesta dataset, respectively. The ablation experiment results show that our proposed method can provide good privacy protection performance with a small sparse factor. Our proposed fraud detection model can significantly improve the fraud detection performance under the joint federated learning framework.

Advanced persistent threat groups are launching extensive attacks on national critical institutions and key infrastructures. Learning contextual information among entities using graph neural network models based on provenance graphs has demonstrated excellent performance in threat detection. However, due to constraints of transmission costs and privacy concerns, existing threat detection methods typically rely solely on audit logs within their own domains, which contain only part of the attacker's behavior. Attackers have also recognized this, so they brazenly carry out attack attempts and expand their impact. The information asymmetry across multiple domains indeed significantly impacts the efficiency and timeliness of APT detection. This paper introduces a novel cross-domain threat detection (CDTD) scheme based on collaborative graph learning for effective and early APT detection. The core idea of CDTD is to foster mutual benefits by utilizing potential cross-domain general knowledge from heterogeneous data and collaborating on precious threat intelligence, which could mitigate information asymmetry. Experimental results demonstrate that the proposed method achieves state-of-the-art detection capabilities and reduces the workload of security analysts. Additionally, the distributed detection scheme significantly decreases system overheads in terms of network bandwidth, memory usage, and execution time, which shows promising potential for broad applications of the proposed scheme.

Workload prediction is a key step in artificial intelligence for IT operations (AIOps) on cloud platforms, enabling proactive application management for performance assurance, cost reduction, and energy optimization. With the popularity of microservice architectures, user requests are now handled collaboratively by multiple service instances, so the workload variation is no longer the individual behavior of each instance but the group behavior of multiple instances. However, existing approaches typically analyze each instance independently and fail to explicitly model group-level workload evolution, leading to suboptimal predictions. To address this issue, we propose MGroup, a workload group behavior-aware multi-instance workload prediction method. First, we define the concept of highly-coupled multi-instance workload group behavior and its evolution, shifting the analytical focus from individuals to groups; second, we propose an adaptive method for identifying and characterizing the workload group behavior based on both static and dynamic correlations, shifting from similarity-based to correlation-based representation; third, we propose a multi-instance parallel prediction neural network that jointly captures local and global workload evolution, shifting from implicit modeling to explicit modeling. Based on this approach, we design a workload prediction system tailored to cloud-native applications. Finally, experimental results on public datasets show that MGroup reduces MAE by 14.62%-21.60% and RMSE by 21.97%-29.27% compared to existing state-of-the-art methods, which provides an effective solution for realizing workload prediction for cloud-native applications.

Microservice architecture is widely adopted in cloud computing for building large-scale distributed systems. However, due to frequent calls between microservices, such systems become increasingly complex and fragile, leading to reduced reliability of the system behavior. Therefore, implementing process discovery and constructing process models for the historical logs of microservice systems becomes crucial to analyzing system behavior. However, existing process discovery methods do not effectively capture the synchronous or asynchronous call relations between microservices at the endpoint level, limiting the ability to analyze the impact of microservice interactions on system performance. This paper proposes a novel process discovery capable of capturing call relations in microservice systems at the endpoint level. First, the call events related to the microservice calls are extracted from the interleaved fragmented raw logs by event serialization to form event sequences. Then, an invocation tree is defined and constructed, and pruning is applied to obtain the subsequence that can be used to derive call relations correctly. Based on the defined event relations, the call relations between endpoints are discovered. Finally, the$\alpha$-based miner is applied to perform the mining and generation of the process model to obtain the defined microservice call net. Experimental results show that, compared to state-of-the-art methods, the proposed approach can accurately and effectively build the microservice system process model.

Fraud detection on dynamic graph (FDDG) is in high demand across many real-world applications, such as financial transaction networks and social networks. A graph neural network (GNN) is an advanced methodology for learning graph data and has been widely adopted in fraud detection tasks. Despite the promising results, research on GNN-based fraud detection faces two critical challenges. First, due to the lack of temporal consistency mining in existing methods, camouflaged fraudsters can easily evade detection by imitating the behavioral patterns of benign users. Second, with constantly evolving fraud patterns and a scarcity of fraud samples, existing methods can easily overfit to current fraud patterns and struggle to identify the evolved and new ones. To address these challenges, we propose CamFD, a semi-supervised camouflage-aware fraud detection model on dynamic graphs. To detect the temporal inconsistency in the behavioral patterns of camouflaged fraudsters, CamFD is equipped with a novel consistency-sensitive contrastive learning (CSCL) module. CSCL discriminatively learns the consistency and inconsistency in users’ behavioral patterns by constructing temporal and structural contrastive pairs. Since modeling the evolving fraud patterns with sparse fraud samples is almost impractical, CamFD concentrates on modeling the relatively stable patterns of extensive benign users with a multivariate Gaussian distribution modeling (MGDM) module. We conduct extensive experiments on a private credit card fraud dataset as well as three public datasets. The experimental results demonstrate that CamFD outperforms ten state-of-the-art (SOTA) baselines across all datasets, with the area under ROC curve (AUC) improvements ranging from 1.03% to 5.32%.

With the rapid advancements in large language models (LLMs), debating tasks, such as argument quality assessment and debate process simulation, have made significant progress. However, existing LLM-based debating systems focus on responding to specific arguments while neglecting objective assessments such as authenticity and logical validity. Furthermore, these systems lack a structured approach to optimize across various dimensions—including evaluation metrics, chain-of-thought (CoT) reasoning, and multi-turn debate refinement—thereby limiting their effectiveness. To address these interconnected challenges, we propose a dual-component framework: (1) InspireScore, a novel evaluation system that establishes a multi-dimensional assessment architecture incorporating four subjective criteria (emotional appeal, argument clarity, argument arrangement, and topic relevance) alongside two objective metrics (fact authenticity and logical validity); and (2) InspireDebate, an optimized debating framework employing a phased optimization approach through CoT reasoning enhancement, multi-dimensional Direct Preference Optimization (DPO), and real-time knowledge grounding via web-based Retrieval Augmented Generation (Web-RAG). Empirical evaluations demonstrate that InspireScore achieves 44% higher correlation with expert judgments compared to existing methods, while InspireDebate shows significant improvements, outperforming baseline models by 57%. Source code is available at https://github.com/fywang12/InspireDebate.

Video Question-Answering (VideoQA) remains challenging in achieving advanced cognitive reasoning due to the uncontrollable and opaque reasoning processes in existing Multimodal Large Language Models (MLLMs). To address this issue, we propose a novel Language-centric Tree Reasoning (LTR) framework that targets on enhancing the reasoning ability of models. In detail, it recursively divides the original question into logically manageable parts and conquers them piece by piece, enhancing the reasoning capabilities and interpretability of existing MLLMs. Specifically, in the first stage, the LTR focuses on language to recursively generate a language-centric logical tree, which gradually breaks down the complex cognitive question into simple perceptual ones and plans the reasoning path through a RAG-based few-shot approach. In the second stage, with the aid of video content, the LTR performs bottom-up logical reasoning within the tree to derive the final answer along with the traceable reasoning path. Experiments across 11 VideoQA benchmarks demonstrate that our LTR framework significantly improves both accuracy and interpretability compared to state-of-the-art MLLMs. To our knowledge, this is the first work to implement a language-centric logical tree to guide MLLM reasoning in VideoQA, paving the way for language-centric video understanding from perception to cognition.

Graph Neural Networks (GNNs) have achieved significant success in various real-world applications, including social networks, finance systems, and traffic management. Recent researches highlight their vulnerability to backdoor attacks in node classification, where GNNs trained on a poisoned graph misclassify a test node only when specific triggers are attached. These studies typically focus on single attack categories and use adaptive trigger generators to create node-specific triggers. However, adaptive trigger generators typically have a simple structure, limited parameters, and lack category-aware graph knowledge, which makes them struggle to handle backdoor attacks across multiple categories as the number of target categories increases. We address this gap by proposing a novel approach for Effective and Unnoticeable Multi-Category (EUMC) graph backdoor attacks, leveraging subgraph from the attacked graph as category-aware triggers to precisely control the target category. To ensure the effectiveness of our method, we construct a Multi-Category Subgraph Triggers Pool (MC-STP) using the subgraphs of the attacked graph as triggers. We then exploit the attachment probability shifts of each subgraph trigger as category-aware priors for target category determination. Moreover, we develop a ``select then attach'' strategy that connects suitable category-aware trigger to attacked nodes for unnoticeability. Extensive experiments across different real-world datasets confirm the efficacy of our method in conducting multi-category graph backdoor attacks on various GNN models and defense strategies.

In the global financial market, assessing bank credit ratings is essential for evaluating financial health, managing risk, and safeguarding systemic stability. While risk can transmit rapidly within the interbank lending network, timely incorporation of the latest financial disclosures to update bank ratings is vital in the swiftly evolving financial markets. However, existing approaches primarily conduct credit rating tasks using end-to-end models trained on historical financial data, thereby overlooking the staggered timing of financial disclosure from banks. Limited excavation of the credit rating records and the temporal distribution shifts existed in different financial periods still pose challenges to improving the accuracy of the credit rating tasks. To address these challenges, in this work we propose a Dual Pairwise pre-training and prompt Tuning framework with Aligned Prototypes (DPTAP) for interbank credit rating, which enables dynamic credit updates. Specifically, the dual pairwise pre-training strategy allows the framework to capture direction and distance discrepancies between rating categories. To alleviate the adverse impact of temporal distribution shifts in quarters, the latest financial features are prompted to dynamically map the patterns of the corresponding banks in the last quarter. Furthermore, we integrate rating guides from two consecutive quarters into a set of aligned prototypes to enhance supervision during the prompt tuning process. We conducted extensive experiments on a real-world bank dataset globally in the latest 8 years. The results demonstrate the superiority of our proposed framework over various competitive models, highlighting its notable capabilities in early warning and risk contagion forecasting.

Spoofing detection in financial trading is crucial, especially for identifying complex behaviors such as conspiracy spoofing. Traditional machine-learning approaches primarily focus on isolated node features, often overlooking the broader context of interconnected nodes. Graph-based techniques, particularly Graph Neural Networks (GNNs), have advanced the field by leveraging relational information effectively. However, in real-world spoofing detection datasets, trading behaviors exhibit dynamic, irregular patterns. Existing spoofing detection methods, though effective in some scenarios, struggle to capture the complexity of dynamic and diverse, evolving inter-node relationships. To address these challenges, we propose a novel framework called the Generative Dynamic Graph Model (GDGM), which models dynamic trading behaviors and the relationships among nodes to learn representations for conspiracy spoofing detection. Specifically, our approach incorporates the generative dynamic latent space to capture the temporal patterns and evolving market conditions. Raw trading data is first converted into time-stamped sequences. Then we model trading behaviors using the neural ordinary differential equations and gated recurrent units, to generate the representation incorporating temporal dynamics of spoofing patterns. Furthermore, pseudo-label generation and heterogeneous aggregation techniques are employed to gather relevant information and enhance the detection performance for conspiratorial spoofing behaviors. Experiments conducted on spoofing detection datasets demonstrate that our approach outperforms state-of-the-art models in detection accuracy. Additionally, our spoofing detection system has been successfully deployed in one of the largest global trading markets, further validating the practical applicability and performance of the proposed method.

Deep neural networks often exhibit sub-optimal performance under covariate and category shifts. Source-Free Domain Adaptation (SFDA) presents a promising solution to this dilemma, yet most SFDA approaches are restricted to closed-set scenarios. In this paper, we explore Source-Free Universal Domain Adaptation (SF-UniDA) aiming to accurately classify "known" data belonging to common categories and segregate them from target-private "unknown" data. We propose a novel Global and Local Clustering (GLC) technique, which comprises an adaptive one-vs-all global clustering algorithm to discern between target classes, complemented by a local k-NN clustering strategy to mitigate negative transfer. Despite the effectiveness, the inherent closed-set source architecture leads to uniform treatment of "unknown" data, impeding the identification of distinct "unknown" categories. To address this, we evolve GLC to GLC++, integrating a contrastive affinity learning strategy. We examine the superiority of GLC and GLC++ across multiple benchmarks and category shift scenarios. Remarkably, in the most challenging open-partial-set scenarios, GLC and GLC++ surpass GATE by 16.8% and 18.9% in H-score on VisDA, respectively. GLC++ enhances the novel category clustering accuracy of GLC by 4.1% in open-set scenarios on Office-Home. Furthermore, the introduced contrastive learning strategy not only enhances GLC but also significantly facilitates existing methodologies.

In federated learning (FL), multiple parties collaborate to train a global model by aggregating their local models while keeping private training sets isolated. One problem hindering effective model aggregation is data heterogeneity. Federated ensemble distillation tackles this problem by using fused local-model knowledge to train the global model rather than directly averaging model parameters. However, most existing methods fuse all knowledge indiscriminately, which makes the global model inherit some data-heterogeneity-caused flaws from local models. While knowledge filtering is a potential coping method, its implementation in FL is challenging due to the lack of public data for knowledge validation. To address this issue, we propose a novel data-free approach (FedKFD) that synthesizes credible labeled data to support knowledge filtering and distillation. Specifically, we construct a prediction capability description to characterize the samples where a local model makes correct predictions. FedKFD explores the intersection of local-model-input space and prediction capability descriptions with a conditional generator to synthesize consensus-labeled proxy data. With these labeled data, we filter for relevant local-model knowledge and further train a robust global model through distillation. The theoretical analysis and extensive experiments demonstrate that our approach achieves improved generalization, superior performance, and compatibility with other FL efforts.

Financial transaction systems have become the critical backbone of modern society, and the sharp increase in fraudulent transactions has become an unavoidable significant topic. Their presence poses a severe threat to financial markets, impacting the health of the economic and social welfare systems of various countries. However, most existing fraud detection methods are limited to detecting individual fraudulent entities within static transaction networks, which are neither suitable for continuously changing dynamic transaction networks nor capable of detecting the increasingly prevalent organized fraud crimes. This paper introduces a novel approach, Parallel Graph Learning with Temporal Stamp Encoding (PGLTSE). On the one hand, it designs a history information module to perform temporal dimension feature learning to adapt to the continuous changes in transaction information in Continuous-Time Dynamic Graphs (CTDG). On the other hand, it designs a gang-aware risk propagation algorithm to infer the risk of organized fraudulent activities in the global transaction relation graph. By simultaneously conducting parallel graph representation learning in both homogeneous global transaction relation graphs and heterogeneous local entity interaction graphs, it aggregates local interaction and global association information for end-to-end training. Extensive experiments on diverse real-world datasets substantiate the superior performance of PGLTSE over existing methods, demonstrating its practical efficacy in detecting complex and evolving fraudulent behaviors in financial networks.

Owing to remarkable advances in 5G and IoT, Mobile Edge Computing (MEC) has been extensively applied to meet the high latency requirements of computation-intensive applications. This paper explores the optimal resource management in MEC involving multiple competitive edge computing servers (ECSs), and presents a Competition-aware differential Pricing and resource Allocation method (ComPA). First, we propose a differential pricing mechanism that jointly analyzes the computing capability and resource usage rate to comprehensively quantify each MU’s use of ECS resources and give a differential per-second price, alleviating the resource underutilization in traditional schemes. Then, to co-optimize ECSs and MUs and ensure all ECSs have a fair chance to increase revenue, a differential pricing-based hierarchy game is developed. Specifically, ECSs and MUs play a Stackelberg game where ECSs price for higher profits and MUs subsequently offload at least-cost. Meanwhile, a non-cooperative game among competing ECSs is formulated. We design a price renewal algorithm that incorporates choice probabilities to find a suboptimal solution iteratively, offering ECSs the most competitive final pricing. MU’s optimal offloading decision is finally derived through convex optimization. Extensive experiments validate the notable superiority of ComPA over other advanced solutions in boosting ECS revenue and MU experience.

This work on model checking of unbounded Petri nets either not really concern the$\omega$-component or only focus on the$\omega$symbols, which may lead to incorrect judgments. This article proposes a model checking approach of$\omega$-independent unbounded Petri nets. First, this approach can ensure the complete state space required for model checking by analyzing the enabled/unenabled marking set of conditionally enabled transition. Second, a comprehensive model checking process of$\omega$-independent unbounded PN is presented, including the generation of extended new modified reachability graph. Third, two theorems are presented to prove that extended new modified reachability graph includes complete reachable markings and sequences of transitions. Finally, the proposed new approach is illustrated through a practical example.

Data sharing facilitates the integration and in-depth exploration of cross-domain data, thereby fostering innovative research and model development. However, privacy leakage emerges as a critical barrier to the sharing and circulating of such data. Existing privacy-preserving technologies face challenges in handling complex scenarios involving multiple participants due to the following reasons: 1) Divergent privacy permission. Data sharing is constrained by various privacy limitations, necessitating the consideration of privacy permissions across different domains, akin to a cross-border process. 2) High collaboration cost. Collaboration among multiple domains to determine the privacy constraint and sharing ways incur additional costs. 3) Large noise magnitude. Traditional privacy techniques to protect the privacy of a single domain using local differential privacy (LDP) may introduce excessive noise, thereby reducing data utility. Drawing inspiration from the cross-border visa issuance process, we present an innovative framework called PriVisa for enabling privacy-preserving data sharing across different domains. It consists of four key modules to overcome the mentioned challenges: the hybrid pattern, optimized sharing path construction, personalized grouping, and LDP-based perturbation. 1) The hybrid pattern for coordination among organizations, considering authentication, privacy constraints, and sharing methods. 2) The optimized sharing path construction using a privacy constraint hierarchy tree to maximize data utility while adhering to privacy requirements. 3) The feature similarity grouping and perturbing mechanism satisfying LDP to protect privacy and optimize data utility. The theoretical and experimental validation confirms PriVisa’s effectiveness in addressing divergent privacy constraints and promoting data utility in cross-domain data sharing.

Anti-money laundering (AML) is crucial to maintaining national financial security. Contemporary AML methods focus on homogeneous mining or unitary money laundering pattern. These methods ignore a characteristic of gang operation in money laundering. Thus, in this paper, we propose a multi-view graph-based hierarchical representation learning method, named MG-HRL, to mine organized money laundering groups. In particular, we extract multi-level representations of transaction subgraphs, including transaction features, user features, structural features, and high-order association features from multiple observational perspectives. To learn the correlation between users, we model transaction networks as heterogeneous information networks (HINs) and design six meta-paths related to money laundering scenarios to mine correlations among users. Combining with correlation representations of users, we propose a heterogeneous hypergraph representation learning method to learn high-order representations of transaction subgraphs. Through hierarchical representation learning, the MG-HRL achieves full exploration of money laundering groups. Finally, we conduct experiments on two public transaction datasets. The result shows that MG-HRL method performs better than other state-of-the-art baselines.

Financial time series prediction is an important and challenging data mining task for quantitative investment. The inherent non-linearity, high noise, and susceptibility to various factors, such as macroeconomic conditions and market sentiment in the stock market, increase the difficulty of prediction. Existing financial industries mainly employ time series models or fundamental analysis methods for prediction. However, these methods fail to effectively capture the complex interrelationships between equity. In recent years, graph neural networks (GNNs), due to their powerful relational modeling capabilities, have been applied to stock prediction. However, with the advances of recent digital power, such as widely-used high-frequency trading techniques, existing graph-based methods still have shortcomings in effectively learning multi-granularity temporal relations as they cannot effectively learn the patterns in different frequencies, e.g., minute-level, daily, weekly, etc. Therefore, in this paper, we propose a multi-granularity graph augmented learning framework for interrelated financial time series forecasting. We first construct a temporal return relationship graph with multi-granularity financial time series, including weekly, daily, and minute-level, to comprehensively capture the dynamic relations of equities, including both medium-term trends and short-term fluctuations. Then, to further augment the node relations, we devise an attentional graph augment module to improve the graph learning with fundamental data, which are jointly optimized in the prediction layer. We conduct extensive empirical studies on multiple datasets from both the Chinese and U.S. stock markets. The results demonstrate that our proposed model consistently outperforms existing baseline methods across four key financial metrics, including ARR, ASR, CR, and IR, thereby validating its effectiveness and superiority. The model has been applied and empirically tested in commercial-grade trading platforms, further demonstrating its efficiency and robustness in real-world trading environments.

As a prominent application of deep neural networks in financial literature, bank credit ratings play a pivotal role in safeguarding global economic stability and preventing crises. In the contemporary financial system, interconnectivity among banks has reached unprecedented levels. However, many existing credit risk models continue to assess each bank independently, resulting in inevitable suboptimal performance. Thus, developing advanced neural networks to model intricate temporal dynamics and interconnected relationships in the banking system is essential for an effective credit rating and risk assessment learning system. To this end, we propose a novel hierarchical fusion transformer for interbank credit rating and risk assessment (HFTCRNet), which includes the long-term temporal transformer (LT3) module, short-term cross-graph transformer (STCGT) module, attentive risk contagion transformer (ARCT) module, and hierarchical fusion transformer (HFT) module to capture the long-term growth trajectories of banks, the short-term interbank network variance, the potential propagation of risks within interbank network, and integrate these information hierarchically. We further develop an interbank credit rating dataset, encompassing quarterly financial data, interbank lending networks, and key indicators such as credit ratings and systemic risk (SRISK) for 4548 banks from 2016Q1 to 2023Q1. Notably, we also adapt the minimum density algorithm to stabilize the interbank loan network over time, aiding in the analysis of long-term and short-term network effects. Our learning system uses semi-supervised learning to handle labels of varying sparsity, integrating credit ratings and SRISK for a comprehensive assessment of individual bank creditworthiness and systemic interbank risk. Extensive experimental results on our interbank dataset show that HFTCRNet not only outperforms all the baselines in terms of credit rating accuracy but also can evaluate the systemic risk within the interbank network. Code will be available at: https://github.com/AI4Risk/HFTCRNet.

Accurately assessing and forecasting bank credit ratings at an early stage is vitally important for a healthy financial environment and sustainable economic development. However, the evaluation process faces challenges due to individual attacks on the rating model. Some participants may provide manipulated information in an attempt to undermine the rating model and secure higher scores, further complicating the evaluation process. Therefore, we propose a novel approach called the preferential selective-aware graph neural network (PSAGNN) model to simultaneously defend against feature and structural nontarget poisoning attacks on Interbank credit ratings. In particular, the model establishes a phased optimization approach combined with biased perturbation and explores the Interbank preferences and scale-free nature of networks, to adaptively prioritize the poisoning training data and simulate a clean graph. Finally, we apply a weighted penalty on the opposition function to optimize the model so that the model can distinguish between attackers. Extensive experiments on our newly collected Interbank quarter dataset and case studies demonstrate the superior performance of our proposed approach in preventing credit rating attacks compared to state-of-the-art baselines.

Currently, due to the different distribution of data for each user, many personalized federated learning (PFL) methods have emerged to meet the personalized needs of different users. However, existing methods have two problems: 1) in the aggregation process, the contribution between the internal layers of the client model is not considered and 2) it is difficult to match the quantitative weight information of both user privacy protection and performance with their qualitative preferences during the training process. Therefore, we first propose a framework for federated aggregation with interlayer personalized contribution named FedIPC, which completes model aggregation based on the contribution of internal layers and improves client model performance. Based on the above framework, we design a multiobjective federated optimization method based on adaptive preference indicators named FedAPI-nondominated sorting genetic algorithm II (NSGA-II). This method can match quantitative weight with qualitative user preferences and adaptively select for Pareto optimal solutions during the optimization process. Extensive experiments on two image datasets and a tabular dataset show that our proposed method not only accelerates model convergence, but also achieves good improvements in model performance. In addition, our proposed method can accurately match the qualitative preferences of users, balancing the performance of the model and privacy protection based on preferences.

The promotion of large-scale applications of reinforcement learning (RL) requires efficient training computation. While existing parallel RL frameworks encompass a variety of RL algorithms and parallelization techniques, the excessively burdensome communication frameworks hinder the attainment of the hardware's limit for final throughput and training effects on a single desktop. In this article, we propose Spreeze, a lightweight parallel framework for RL that efficiently utilizes a single desktop hardware resource to approach the throughput limit. We asynchronously parallelize the experience sampling, network update, performance evaluation, and visualization operations, and employ multiple efficient data transmission techniques to transfer various types of data between processes. The framework can automatically adjust the parallelization hyperparameters based on the computing ability of the hardware device in order to perform efficient large-batch updates. Based on the characteristics of the “Actor-Critic” RL algorithm, our framework uses dual GPUs to independently update the network of actors and critics in order to further improve throughput. Simulation results show that our framework can achieve up to 15,000 Hz experience sampling and 370,000 Hz network update frame rate using only a personal desktop computer, which is an order of magnitude higher than other mainstream parallel RL frameworks, resulting in a 73% reduction of training time. Our work on fully utilizing the hardware resources of a single desktop computer is fundamental to enabling efficient large-scale distributed RL training.

Medical insurance fraud has always been a crucial challenge in the field of healthcare industry. Existing fraud detection models mostly focus on offline learning scenes. However, fraud patterns are constantly evolving, making it difficult for models trained on past data to detect newly emerging fraud patterns, posing a severe challenge in medical fraud detection. Moreover, current incremental learning models are mostly designed to address catastrophic forgetting, but often exhibit suboptimal performance in fraud detection. To address this challenge, this paper proposes an innovative online learning method for medical insurance fraud detection, named POCL. This method combines contrastive learning pre-training with online updating strategies. In the pre-training stage, we leverage contrastive learning pre-training to learn on historical data, enabling deep feature learning and obtaining rich risk representations. In the online learning stage, we adopt a Temporal Memory Aware Synapses online updating strategy, allowing the model to perform incremental learning and optimization based on continuously emerging new data. This ensures timely adaptation to fraud patterns and reduces forgetting of past knowledge. Our model undergoes extensive experiments and evaluations on real-world insurance fraud datasets. The results demonstrate our model has significant advantages in accuracy compared to the state-of-the-art baseline methods, while also exhibiting lower running time and space consumption. Our sources are released at https://github.com/finint/POCL.

Multi-constraint offline reinforcement learning (RL) promises to learn policies that satisfy both cumulative and state- wise costs from offline datasets. This arrangement provides an effective approach for the widespread appli-cation of RL in high-risk scenarios where both cumulative and state-wise costs need to be considered simulta-neously. However, previously constrained offline RL algorithms are primarily designed to handle single-constraint problems related to cumulative cost, which faces challenges when addressing multi-constraint tasks that involve both cumulative and state-wise costs. In this work, we pro-pose a novel Primal policy Optimization with Conservative Estimation algorithm (POCE) to address the problem of multi-constraint offline RL. Concretely, we reframe the ob-jective of multi-constraint offline RL by introducing the con-cept of Maximum Markov Decision Processes (MMDP). Subsequently, we present a primal policy optimization al-gorithm to confront the multi-constraint problems, which improves the stability and convergence speed of model training. Furthermore, we propose a conditional Bell-man operator to estimate cumulative and state-wise Q-values, reducing the extrapolation error caused by out-of-distribution (OOD) actions. Finally, extensive experiments demonstrate that the POCE algorithm achieves competitive performance across multiple experimental tasks, particu-larly outperforming baseline algorithms in terms of safety. Our code is available at github. POCE.

Deep learning has achieved remarkable progress in various applications, heightening the importance of safeguarding the intellectual property (IP) of well-trained models. It entails not only authorizing usage but also ensuring the deployment of models in authorized data domains, i.e., making models exclusive to certain target domains. Previous methods necessitate concurrent access to source training data and target unauthorized data when performing IP protection, making them risky and inefficient for decentralized private data. In this paper, we target a practical setting where only a well-trained source model is available and investigate how we can realize IP protection. To achieve this, we propose a novel MAsk Pruning (MAP) framework. MAP stems from an intuitive hypothesis, i.e., there are target-related parameters in a well-trained model, locating and pruning them is the key to IP protection. Technically, MAP freezes the source model and learns a target-specific binary mask to prevent unauthorized data usage while minimizing performance degradation on authorized data. Moreover, we introduce a new metric aimed at achieving a better balance between source and target performance degradation. To verify the effectiveness and versatility, we have evaluated MAP in a variety of scenarios, including vanilla source-available, practical source-free, and challenging data-free. Extensive experiments indicate that MAP yields new state-of-the-art performance. Code will be available at https://github.com/ispc-lab/MAP.

Universal Domain Adaptation (UniDA) targets knowledge transfer in the presence of both covariate and label shifts. Recently, Source-free Universal Domain Adaptation (SF-UniDA) has emerged to achieve UniDA without access to source data, which tends to be more practical due to data protection policies. The main challenge lies in determining whether covariate-shifted samples belong to target-private unknown categories. Existing methods tackle this either through hand-crafted thresholding or by developing time-consuming iterative clustering strategies. In this paper, we propose a new idea of LEArning Decomposition (LEAD), which decouples features into source-known and-unknown components to identify target-private data. Technically, LEAD initially leverages the or-thogonal decomposition analysis for feature decomposition. Then, LEAD builds instance-level decision boundaries to adaptively identify target-private data. Extensive experiments across various UniDA scenarios have demonstrated the effectiveness and superiority of LEAD. Notably, in the OPDA scenario on VisDA dataset, LEAD outperforms GLC by 3.5% overall H-score and reduces 75% time to derive pseudo-labeling decision boundaries. Besides, LEAD is also appealing in that it is complementary to most existing methods. The code is available at https://github.com/ispc-lab/LEAD.

Although neural radiance fields (NeRFs) have achieved triumphs in image novel view synthesis (NVS), LiDAR NVS remains largely unexplored. Previous LiDAR NVS methods employ a simple shift from image NVS methods while ignoring the dynamic nature and the large-scale reconstruction problem of LiDAR point clouds. In light of this, we propose LiDAR4D, a differentiable LiDAR-only frameworkfor novel space-time LiDAR view synthesis. In consideration of the sparsity and large-scale characteristics, we design a 4D hybrid representation combined with multi-planar and grid features to achieve effective reconstruction in a coarse-to-fine manner. Furthermore, we introduce geometric constraints derivedfrom point clouds to improve temporal consistency. For the realistic synthesis of LiDAR point clouds, we incorporate the global optimization of ray-drop prob-ability to preserve cross-region patterns. Extensive experiments on KITTI-360 and NuScenes datasets demonstrate the superiority of our method in accomplishing geometry-aware and time-consistent dynamic reconstruction. Codes are available at https://github.com/ispc-labILiDAR4D.

Collaborative perception is dedicated to tackling the constraints of single-agent perception, such as occlusions, based on the multiple agents' multi-view sensor inputs. However, most existing works assume an ideal condition that all agents' multi-view cameras are continuously available. In reality, cameras may be highly noisy, obscured or even failed during the collaboration. In this work, we introduce a new robust camera-insensitivity problem: how to overcome the issues caused by the failed camera perspectives, while stabilizing high collaborative performance with low calibration cost? To address above problems, we propose RCDN, a Robust Camera-insensitivity collaborative perception with a novel Dynamic feature-based 3D Neural modeling mechanism. The key intuition of RCDN is to construct collaborative neural rendering field representations to recover failed perceptual messages sent by multiple agents. To better model collaborative neural rendering field, RCDN first establishes a geometry BEV feature based time-invariant static field with other agents via fast hash grid modeling. Based on the static background field, the proposed time-varying dynamic field can model corresponding motion vector for foregrounds with appropriate positions. To validate RCDN, we create OPV2V-N, a new large-scale dataset with manual labelling under different camera failed scenarios. Extensive experiments conducted on OPV2V-N show that RCDN can be ported to other baselines and improve their robustness in extreme camera-insensitivity setting. Our code and datasets will be available soon.

Although recent efforts have extended Neural Radiance Field (NeRF) into LiDAR point cloud synthesis, the majority of existing works exhibit a strong dependence on precomputed poses. However, point cloud registration methods struggle to achieve precise global pose estimation, whereas previous pose-free NeRFs overlook geometric consistency in global reconstruction. In light of this, we explore the geometric insights of point clouds, which provide explicit registration priors for reconstruction. Based on this, we propose Geometry guided Neural LiDAR Fields (GeoNLF), a hybrid framework performing alternately global neural reconstruction and pure geometric pose optimization. Furthermore, NeRFs tend to overfit individual frames and easily get stuck in local minima under sparse-view inputs. To tackle this issue, we develop a selective-reweighting strategy and introduce geometric constraints for robust optimization. Extensive experiments on NuScenes and KITTI-360 datasets demonstrate the superiority of GeoNLF in both novel view synthesis and multi-view registration of low-frequency large-scale point clouds.

Heterogeneous data silos hinder the application of deep learning in the Internet of Things. As a dealing scheme, personalized federated learning (pFL) distributedly customizes multiple local models for these silos. Most pFL methods directly use a global model to assist the local model optimization ignoring the performance drop caused by irrelevant or misleading global-model knowledge. To address this, we propose an adaptive knowledge recomposition approach (FedAKR), which refines relevant and correctly leading knowledge from the global model and recomposes it into the local model to promote personalization. Specifically, FedAKR provides a discrete wavelet transform-based method to recompose different kinds of knowledge in the same representation space. Facilitated by this common space, we introduce an enriched local optimization objective to establish a causal relationship between the refined global-model knowledge and recomposed local-model knowledge. The relationship guides effective and efficient knowledge refinement, thereby promoting personalization. Besides, we provide the theoretical proof of convergence for our novel pFL approach. Extensive experiments demonstrate that FedAKR achieves interpretable improvements, higher performance over 12 state-of-the-art methods, and the potential to further integrate pretrained large models.

Graph convolutional networks (GCNs) can quickly and accurately learn graph representations and have shown powerful performance in many graph learning domains. Despite their effectiveness, neighborhood awareness remains essential and challenging for GCNs. Existing methods usually perform neighborhood-aware steps only from the node or hop level, which leads to a lack of capability to learn the neighborhood information of nodes from both global and local perspectives. Moreover, most methods learn the nodes' neighborhood information from a single view, ignoring the importance of multiple views. To address the above issues, we propose a multi-view adaptive neighborhood-aware approach to learn graph representations efficiently. Specifically, we propose three random feature masking variants to perturb some neighbors' information to promote the robustness of graph convolution operators at node-level neighborhood awareness and exploit the attention mechanism to select important neighbors from the hop level adaptively. We also utilize the multi-channel technique and introduce a proposed multi-view loss to perceive neighborhood information from multiple perspectives. Extensive experiments show that our method can better obtain graph representation and has high accuracy.

Many studies have achieved excellent performance in analyzing graph-structured data. However, learning graph-level representations for graph classification is still a challenging task. Existing graph classification methods usually pay less attention to the fusion of node features and ignore the effects of different-hop neighborhoods on nodes in the graph convolution process. Moreover, they discard some nodes directly during the graph pooling process, resulting in the loss of graph information. To tackle these issues, we propose a new Graph Multi-Convolution and Attention Pooling based graph classification method (GMCAP). Specifically, the designed Graph Multi-Convolution (GMConv) layer explicitly fuses node features learned from different perspectives. The proposed weight-based aggregation module combines the outputs of all GMConv layers, for adaptively exploiting the information over different-hop neighborhoods to generate informative node representations. Furthermore, the designed Local information and Global Attention based Pooling (LGAPool) utilizes the local information of a graph to select several important nodes and aggregates the information of unselected nodes to the selected ones by a global attention mechanism when reconstructing a pooled graph, thus effectively reducing the loss of graph information. Extensive experiments show that GMCAP outperforms the state-of-the-art methods on graph classification tasks, demonstrating that GMCAP can learn graph-level representations effectively.

Abstract Software defect prediction (SDP) aims to build an effective prediction model for historical defect data from software repositories by some specialized techniques or algorithms, and predict the defect proneness of new software modules. Nevertheless, the complex internal intrinsic structure hidden behind the defect data makes it challenging for the built prediction model to capture the most expressive defect feature representations, and largely limits the SDP performance. Fortunately, artificial intelligence is interacting closely with humans and provides powerful intelligent technical support for addressing these SDP issues. In this article, we propose a robust intelligent SDP model called IMDAC based on deep learning and soft computing techniques. This model has three main advantages: (1) an effective deep generative network—InfoGAN (information maximizing GANs) is employed to conduct data augmentation, namely generating sufficient defect instances and achieving defect class balance simultaneously. (2) Select the fewest representative feature subset for the minimum error via an advanced multi‐objective optimization approach—MSEA (multi‐stage evolutionary algorithm). (3) Build a powerful end‐to‐end deep defect predictor by hybrid deep learning techniques—DAE (Denoising AutoEncoder) and CNN (convolutional neural network), which can not only reconstruct a clean “repaired” input with strong robustness and generalization capabilities via DAE, but also learn the abstract deep semantic features with strong discriminating capability via CNN. Experimental results verify the superiority and robustness of the IMDAC model across 15 software projects.

Unbounded Petri nets (UPNs) can describe and analyze discrete event systems with infinite states (DESIS). Due to the infinite state space and the combination explosion problem, the reachability analysis of UPNs is an NP-Hard problem. The existing reachability analysis methods cannot achieve an accurate result at reasonable costs (computational time and space) due to the finite reachability tree with$\omega$-numbers. Based on the idea of approximating infinite space with finite states, given some limited reachable markings of a UPN, we propose a method that can quantitatively solve the UPN’s reachability problem with machine learning. Firstly, we define the probabilistic reachability of markings and transform the UPN’s reachability problem into the prediction problem of markings. The proposed method based on positive and unlabeled learning (PUL) and bagging trains a classifier to predict the probabilistic reachability of unknown markings. Finally, to predict the markings outside the positive sample set and unlabeled sample set, an iterative strategy is designed to update the classifier. Based on seven general UPNs, the results of the experiments show that the proposed method has a good performance in the accuracy and time consumption for the UPN’s reachability problem.Note to Practitioners—In discrete event systems, the reachability problem mainly studies reachable states of the system and the relationship between states, which is the basis of the system’s states, behaviors, attributes and performance analysis. For discrete event systems with infinite states, it is hard to analyze the reachable relationship between states within a finite time due to the infinite state space and the combination explosion problem. The main motivation of the paper is to propose a method that can predict the reachable relationship between the states with a probability value within a finite time. By machine learning algorithms, the method learns the feature information of the known reachable states. The reachability of unknown states in the infinite state space can be predicted approximately. The proposed approximation method can be applied to analyze the reachability properties of general discrete event systems with infinite states, such as checking whether a fault occurs in operating systems, whether a message is delivered in communication and so on.

The innovations in technologies have led to the emergence of digital finance such as online payment, online insurance, online lending, and supply chain finance. Digital finance has greatly facilitated people’s lives, accelerated the circulation of capital in various fields, and enhanced the vitality of financial markets. However, it exposes many increasing risks and hidden dangers such as stock volatility, trading fraud, credit card fraud, and privacy leakage[1],[2],[3],[4],[5],[6],[7]. How to effectively calculate, control, manage, and utilize financial big data and make full use of artificial intelligence technology to ensure financial security is an important research question. Solving it faces many challenges. These challenges not only include the complexity of data and computation but also the effectiveness of intelligent optimization algorithms and ways to deal with human behaviors and social environments[8],[9].

Internet of Vehicles (IoV) improves traffic safety and efficiency by wireless communications among vehicles and infrastructures. To ensure secure communications in IoV, the problem of vehicle identity security must be solved before deployment. In this article, we propose a quick-response behavior-based vehicle identity verification method, called VLOG, for solving identity theft in IoV. This method is based on the idea of a vehicle usually having relatively stable traveling habit/behaivor. If we detect unusual behavior, the vehicle's identity may be stolen. VLOG captures vehicles’ latent behavior models from local and global two aspects, and further merges local and global models into a comprehensive behavior-based identity verification model. In the local part, we give a 2-D Gaussian model to fit the behavior data. In the global part, we learn vehicles’ traveling preferences under secure multiparty computation framework with considering the behavior volatility. The results of experiments based on a real-world vehicular trace dataset show the best performance of VLOG in terms of accuracy, F1 score, and cost. Meanwhile, VLOG also performs well in the area under the curve and precision-recall curve. Besides, since our model is preprepared, when a vehicle is required to be detected, the verification response time is short.

With the rapid development of Internet industry, the learning methods based on data stream have attracted more and more attention owing to their great application value in many industries such as banking, insurance, and telecom industry. In the process of learning from stream data, one of the most significant challenges is how to adapt to the so-called concept drift which means that the data stream distribution changes over time in unpredictable ways. To deal with the problem, some methods have been put forward. However, most of them pay more attention to the whole entity of a given model and overlook the impact of local data on the model. In this article, we propose a novel ensemble algorithm to overcome the problem of ignoring local samples in previous methods, namely, the fit prediction algorithm based on local optimum (FPLO) which uses the information of local data to fit (predict) a concept drift. Furthermore, an adaptive method based on the so-called concept changing rate is proposed to choose those classifiers of suitable sizes to overcome the problem of selecting too much or too little historical drift information in previous methods and to make the prediction more accurate. The experimental results on nine synthetic stream datasets and eight real-world stream datasets which all have the concept drift problem show that our FPLO is able to tackle the problem more effectively in comparison to other state-of-the-art methods.

Industries, such as manufacturing, are accelerating their embrace of the metaverse to achieve higher productivity, especially in complex industrial scheduling. In view of the growing parking challenges in large cities, high-density vehicle spatial scheduling is one of the potential solutions. Stack-based parking lots utilize parking robots to densely park vehicles in the vertical stacks like container stacking, which greatly reduces the aisle area in the parking lot, but requires complex scheduling algorithms to park and take out the vehicles. The existing high-density parking (HDP) scheduling algorithms are mainly heuristic methods, which only contain simple logic and are difficult to utilize information effectively. We propose a hybrid residual multiexpert (HIRE) reinforcement learning (RL) approach, a method for interactive learning in the digital industrial metaverse, which efficiently solves the HDP batch space scheduling problem. In our proposed framework, each heuristic scheduling method is considered as an expert. The neural network trained by RL assigns the expert strategy according to the current parking lot state. Furthermore, to avoid being limited by heuristic expert performance, the proposed hierarchical network framework also sets up a residual output channel. Experiments show that our proposed algorithm outperforms various advanced heuristic methods and the end-to-end RL method in the number of vehicle maneuvers, and has good robustness to the parking lot size and the estimation accuracy of vehicle exit time. We believe that the proposed HIRE RL method can be effectively and conveniently applied to practical application scenarios, which can be regarded as a key step for RL to enter the practical application stage of the industrial metaverse.

Anti-fraud engineering for online credit loan (OCL) platforms is getting more challenging due to the developing specialization of gang fraud. Associations are critical features referring to assessing the credibility of loan applications for OCL fraud prediction. State-of-the-art solutions employ graph-based methods to mine hidden associations among loan applications effectively. They perform well based on the information asymmetry which is guaranteed by the huge advantage of platforms over fraudsters in terms of data quantity and quality at their disposal. The inherent difficulty that can be foreseen is thedata isolationcaused by mistrust between multiple platforms and data control legislations for privacy preservation. To maintain the advantage owned by the platforms, we design a privacy-preserving distributed graph learning framework that ensures critical association repairs by merging parameter sharing and data sharing. Specially, we propose theassociation reconstruction mechanism(ARM) that consists of the devised exploration, processing, transmission and utilization schemes to realize data sharing. For parameter sharing, we design a hybrid encryption technique to protect privacy during collaboratively learning graph neural network (GNN) models among different financial client platforms. We conduct the experiments over real-life data from large financial platforms. The results demonstrate the effectiveness and efficiency of our proposed methods.

Nowadays, a lot of studies unite multiple organizations to form an anti-fraud alliance to detect fraudulent transactions better using federated boosting tree algorithms. However, there are two challenges when building the federated boosting tree-based fraud detection model. First, the vertical federated learning (VFL) framework is not enough for the transaction fraud detection task because there are various participants in the same field (e.g., different banks) who cannot share data with others freely. And a lot of local data would be discarded due to the private set intersection process under the VFL framework. Second, there are still many privacy threats that can infer the data sequence information of the participants. Once the attackers illegally obtain the data sequence information, they can infer the raw data of the victims based on the data distribution. Specially, the instance spaces and score lists would also be maliciously exploited to launch a new data sequence inference attack that is currently indefensible. In response to the above challenges, we first propose a sequence inference attack that is the first work showing the vulnerabilities regarding the instance spaces and score lists. Then, we propose a federated boosting tree-based fraud detection method against (A) sequence (S) inference (I) attacks (A), named ASIA. ASIA method can combine the horizontal federated learning (HFL) framework with the VFL framework to better detect fraudulent transactions while defending against sequence inference attacks. Finally, we evaluate SIS attack and ASIA method in experiments based on two public fraud detection datasets: European Credit Card (ECC) and IEEE-CIS Fraud Detection (Vesta). The experimental results show that the sequences of the participants are at a high risk of being leaked when suffering SIS attack. Moreover, compared with several widely used federated boosting tree methods, ASIA method can significantly improve privacy-preserving performance without sacrificing fraud detection accuracy.

The possibility for public attributes to disclose private information has caused widespread concern. Traditional privacy-preserving approaches have two limitations: 1) Approaches based on data anonymization or distortion often lead to poor utility-privacy trade-offs, and 2) approaches based on data encryption face heavy computational costs. These problems have prompted calls for an effective privacy-preserving framework that provides adequate privacy guarantees while maintaining good data utility. Inspired by denoising autoencoders, in this paper, we regard the information about privacy attributes contained in the public attributes as a kind of noise and design an ex ante privacy-preserving model called the Mutual Information Autoencoder (MIAE), which reconstructs the loss function of the original autoencoder by combining reconstruction errors and mutual information, and we introduce a trade-off coefficient to achieve utility-privacy trade-offs. To elucidate the superiority of the proposed model, we consider utility-privacy trade-offs with the expected distortion function as a metric of data utility and the joint mutual information as a metric of privacy disclosure, and then, we construct a convex optimization problem with multiple constraints based on rate-distortion theory. From an information theory perspective, we provide a lower bound for privacy disclosure with utility guarantees. Elaborate experiments over a real-world dataset reveal that as the level of expected distortion increases, the achievable bound obtained by MIAE exhibits a trend similar to that of the information-theoretic bound. When the expected distortion surpasses 2.2, the achievable bound obtained by MIAE also converges to 0, and the maximum gap between the achievable bound obtained by MIAE and the information-theoretic bound is no more than 1.4. Compared to existing models, MIAE can provide a tighter achievable bound and achieve good utility-privacy trade-offs.

Graph neural networks (GNNs) are playing exciting roles in the application scenarios where features are hidden in information associations. Fraud prediction of online credit loan services (OCLSs) is such a typical scenario. But it has another rather critical challenge, i.e., the scarcity of data labels. Fortunately, GNNs can also cope with this problem due to their good ability of semi-supervised learning by mining structure and feature information within graphs. Nevertheless, the gain of internal information is often too limited to help GNNs handle the extreme deficiency of labels with high performance beyond the basic requirement of fraud prediction in OCLSs. Therefore, adding labels from the experts, such as manually adding labels through rules, has become a logical practice. However, the existing rule engines for OCLSs have the confliction problem among continuously accumulated rules. To address this issue, we propose a Snorkel-based Semi-Supervised GNN (S3GNN). Under S3GNN, we specially design an upgraded version of the rule engines, called Graph-Oriented Snorkel (GOS), a graph-specific extension of Snorkel, a widely used weakly supervised learning framework, to design rules by subject matter experts (SMEs) and resolve confliction. In particular, in the graph of an anti-fraud scenario, each node pair may have multiple different types of edges, so we propose the Multiple Edge-Types Based Attention mechanism. In general, for the heterogeneous information and multiple relations in the graph, we first obtain the embedding of applicant nodes by aggregating the representation of attribute nodes, and then use the attention mechanism to aggregate neighbor nodes on multiple meta-paths to get ultimate applicant node embedding. We conduct experiments over the real-life data of a large financial platform. The results demonstrate that S3GNN can outperform the state-of-the-art methods, including the method of pilot platform.

Graph convolutional networks (GCNs) have shown superior performance on graph classification tasks, and their structure can be considered as an encoder-decoder pair. However, most existing methods lack the comprehensive consideration of global and local in decoding, resulting in the loss of global information or ignoring some local information of large graphs. And the commonly used cross-entropy loss is essentially an encoder-decoder global loss, which cannot supervise the training states of the two local components (encoder and decoder). We propose a multichannel convolutional decoding network (MCCD) to solve the above-mentioned problems. MCCD first adopts a multichannel GCN encoder, which has better generalization than a single-channel GCN encoder since different channels can extract graph information from different perspectives. Then, we propose a novel decoder with a global-to-local learning pattern to decode graph information, and this decoder can better extract global and local information. We also introduce a balanced regularization loss to supervise the training states of the encoder and decoder so that they are sufficiently trained. Experiments on standard datasets demonstrate the effectiveness of our MCCD in terms of accuracy, runtime, and computational complexity.

Credit card fraud detection is a challenging task since fraudulent actions are hidden in massive legitimate behaviors. This work aims to learn a new representation for each transaction record based on the historical transactions of users in order to capture fraudulent patterns accurately and, thus, automatically detect a fraudulent transaction. We propose a novel model by improving long short-term memory with a time-aware gate that can capture the behavioral changes caused by consecutive transactions of users. A current-historical attention module is designed to build up connections between current and historical transactional behaviors, which enables the model to capture behavioral periodicity. An interaction module is designed to learn comprehensive and rational behavioral representations. To validate the effectiveness of the learned behavioral representations, experiments are conducted on a large real-world transaction dataset provided to us by a financial company in China, as well as a public dataset. Experimental results and the visualization of the learned representations illustrate that our method delivers a clear distinction between legitimate behaviors and fraudulent ones, and achieves better fraud detection performance compared with the state-of-the-art methods.

The explosive development of the Internet of Things (IOT) has stimulated the sudden and dynamic demand pattern of potential Service Function Chain (SFC) traffic. The virtual network built based on the requirements of Multiple SFCs (Multi-SFCs), sharing the resource, should provide better Quality of Experience (QoE) for Multi-SFCs according to the real-time network status. However, simultaneous Multi-SFCs in the routing process compete for the VNF resources of the same node and preempt the same link. The instability of QoE caused by node invalidation and link faults is highlighted. Therefore, to provide a more efficient and stable service to Multi-SFCs, quantitative QoE utility and stability models and cooperative Multi-SFCs path allocation are needed. Considering the limited resources, this paper designs a stable QoE-aware Multi-SFCs Cooperative Routing Mechanism (CRM) based on Deep Reinforcement Learning (DRL). Firstly, we model the problem of Multi-SFCs cooperative routing aiming at optimizing diverse QoEs of utility, stability and delay, and formulate it as a multi-objective optimization problem. Moreover, the QoE stable queueing model based on node and link availability is formulated by Lyapunov drift, which can potentially mitigate the QoE fluctuations. Then we associate triple Dueling Deep Q-Networks (DDQNs) and propose a cooperative computing multi-task DRL to obtain the optimal path allocation policy. Experiments are conducted to verify the effectiveness and the results show that our method outperforms its peers on QoE, delay, throughput and stability.

Serverless computing is a popular cloud computing model that offers on-demand resource allocation and pay-as-you-go application execution. However, there are still challenges in allocating resources for workflow applications: inaccurate and inefficient resource estimation, high-latency inter-function communication, and long server readiness time. Therefore, we propose the heterogeneity-aware Proactive serverLess wOrkflow Elastic Allocation method (PLOEA) to address these issues and optimize infrastructure costs for cloud service providers (CSPs) while meeting the diverse needs of developers. Specifically, we propose a resource configuration estimation method for heterogeneous workflow applications that builds an ensemble multi-task expert classifier to analyze individual and common resource usage patterns, ensuring estimation accuracy and efficiency. Further, we propose a group allocation strategy for multiple applications that optimizes the spatiotemporal distribution of instances by considering the allocation urgency, communication affinity between functions, and the multi-core architecture of servers. Furthermore, we present a proactive server elastic scaling method that senses workload features, including workload level, trend, and magnitude changes, and combines them with CSP's attention differences to guide the server scaling size. Finally, experiments based on public datasets prove that PLOEA provides better service quality and cost efficiency than existing methods.

Each activity in the service workflow interacts with services as required to meet complex business needs and quickly adapt to market changes. The design of each activity's input/output interface parameters influences whether it can successfully map to appropriate and interactive services. In practice, suitable activity interface parameters should possess 3-features:$realism$,$relevance$, and$compatibility$, as popular parameters originating from the real world and closely related to activity semantics are apt to match user-expected services. However, existing research requires expert specification or ontology-based inference, resulting in outdated, inconsistent parameters that lack necessary elements, making it challenging to match expected services. Therefore, we propose an automated method combining Transformer and weighted HITS to recommend interface parameters with 3-features on activity function requirement. It filters similar Endpoints (EPs) based on the activity's semantics by supervised Transformer-based learning of multi-domain APIs and unsupervised EPs matching. Next, a node-weighted heterogeneous graph is built based on similar EPs and their interface parameter relationships. We then apply a node-weighted HITS to explore mutual gain relationships within the graph and calculate parameter compatibilities. Finally, a top-$k$non-redundant compatible parameter list and corresponding different formats are recommended for the activity. The method's effectiveness and efficiency are verified using a real API service dataset from RapidAPI.

Monolithic architectures are becoming increasingly difficult to cope with complex applications, and microservice architectures, which offer flexibility and logical independence in development and maintenance, are the new choice for companies and developers. Migrating a legacy monolithic architecture application to a microservice architecture rather than building it from scratch is considered an easy way to use it. To ensure that the migrated microservice applications can take advantage of their benefits, we need to propose a reasonable and effective microservice extraction method. Considering the single responsibility principle in the microservice design principle, most existing microservice extraction methods only pursue the high logical independence of the extraction results and pay little attention to whether the extraction results have good performance. Applications need to perform well, and studies have shown that poor microservice extraction schemes can negatively impact the performance of the migrated application. As a result, when extracting, we should also consider the performance of the results. A few studies consider the performance of extraction results, but only in terms of a few factors affecting performance, such as network overhead, rather than considering all factors affecting performance comprehensively, which leads to an inaccurate evaluation of performance. Therefore, oriented toward the most widely used managed languages today, we propose an effective Microservice Extraction method based on a Comprehensive Evaluation of logical independence and performance (MECE). Firstly, we propose a workflow-based approach to evaluate the performance of microservice extraction results by considering multiple influencing factors, focusing on the management cost ignored in existing studies, and designing an effective management cost evaluation model. After that, we propose a meta-heuristic search-based algorithm to obtain feasible microservice extraction results. In experiments based on actual deployments, the extraction results of the MECE method obtained a performance improvement of up to 46.15% without significant loss of logical independence compared to existing methods, which verifies the effectiveness of the method.

The liveness of Petri nets (PNs) means that every event can occur in any state, establishing a close relationship with the deadlock-free property of existing systems. Due to the problem of state space explosion and the infinite state space of unbounded PNs (UPNs), the time complexity and space complexity of the liveness decision are difficult to give accurate measures; at least, they are both NP-hard. Except for some particular subclasses of UPNs, there has not been an accurate method to decide the liveness of generalized UPNs. Thus, a liveness decision method from a machine learning perspective is proposed to predict probability values about UPNs’ liveness within a finite time. The method aims to learn the feature information on UPNs by deep neural networks and establish the mapping relationship between the UPNs and the liveness. First, the concept of approximating infinite space with finite states is applied to generate reachability graphs at different moments, following the firing rules of a UPN. Then, the graph convolutional network (GCN)-based reachability graph feature representation module and the gated recurrent unit (GRU)-based UPN feature representation module are designed to map the reachability graphs at different moments into the low-dimensional feature space. And the feature vector that can characterize the UPN’s liveness is obtained to decide the liveness probabilistically. Finally, three datasets, including 50000 samples, are constructed. Based on these datasets and some case studies, the experimental results validate the method’s ability to make liveness decisions for UPNs, demonstrating its strong performance in terms of effectiveness and generalization.

The networked-loan is major financing support for Micro, Small and Medium-sized Enterprises (MSMEs) in some developing countries. But external shocks may weaken the financial networks' robustness; an accidental default may spread across the network and collapse the whole network. Thus, predicting the critical firms in networked-loans to stem contagion risk and prevent potential systemic financial crises is of crucial significance to the long-term health of inclusive finance and sustainable economic development. Existing approaches in the banking industry dismiss the contagion risk across loan networks and need extensive knowledge with sophisticated financial expertise. Regarding the issues, we propose a novel approach to predict critical firms for stemming contagion risk in the bank industry with deep reinforcement learning integrated with high-order graph message-passing networks. We demonstrate that our approach outperforms the state-of-the-art baselines significantly on the dataset from a large commercial bank. Moreover, we also conducted empirical studies on the real-world loan dataset for risk mitigation. The proposed approach enables financial regulators and risk managers to better track and understands contagion and systemic risk in networked-loans. The superior performance also represents a paradigm shift in addressing the modern challenges in financing support of MSMEs and sustainable economic development.

Deep neural networks (DNNs) usually fail to generalize well to outside of distribution (OOD) data, especially in the extreme case of single domain generalization (single-DG) that transfers DNNs from single domain to multiple unseen domains. Existing single-DG techniques commonly devise various data-augmentation algorithms, and remould the multi-source domain generalization methodology to learn domain-generalized (semantic) features. Nevertheless, these methods are typically modality-specific, thereby being only applicable to one single modality (e.g., image). In contrast, we target a versatile Modality-Agnostic Debiasing (MAD) framework for single-DG, that enables generalization for different modalities. Technically, MAD introduces a novel two-branch classifier: a biased-branch encourages the classifier to identify the domain-specific (superficial) features, and a general-branch captures domain-generalized features based on the knowledge from biased-branch. Our MAD is appealing in view that it is pluggable to most single-DG models. We validate the superiority of our MAD in a variety of single-DG scenarios with different modalities, including recognition on 1D texts, 2D images, 3D point clouds, and semantic segmentation on 2D images. More remarkably, for recognition on 3D point clouds and semantic segmentation on 2D images, MAD improves DSU by 2.82% and 1.5% in accuracy and mIOU.

Deep neural networks (DNNs) often perform poorly in the presence of domain shift and category shift. How to upcycle DNNs and adapt them to the target task remains an important open problem. Unsupervised Domain Adaptation (UDA), especially recently proposed Source-free Domain Adaptation (SFDA), has become a promising technology to address this issue. Nevertheless, existing SFDA methods require that the source domain and target domain share the same label space, consequently being only applicable to the vanilla closed-set setting. In this paper, we take one step further and explore the Source-free Universal Domain Adaptation (SF-UniDA). The goal is to identify “known” data samples under both domain and category shift, and reject those “unknown” data samples (not present in source classes), with only the knowledge from standard pre-trained source model. To this end, we introduce an innovative global and local clustering learning technique (GLC). Specifically, we design a novel, adaptive one-vs-all global clustering algorithm to achieve the distinction across different target classes and introduce a local k-NN clustering strategy to alleviate negative transfer. We examine the superiority of our GLC on multiple benchmarks with different category shift scenarios, including partial-set, open-set, and open-partial-set DA. Remarkably, in the most challenging open-partial-set DA scenario, GLC outperforms UMAD by 14.8% on the VisDA benchmark. The code is available at https://github.com/ispc-lab/GLC.

In recent years, there has been a significant increase in focus on the interpolation task of computer vision. Despite the tremendous advancement of video interpolation, point cloud interpolation remains insufficiently explored. Meanwhile, the existence of numerous nonlinear large motions in real-world scenarios makes the point cloud interpolation task more challenging. In light of these issues, we present NeuralPCI: an end-to-end 4D spatiotemporal Neural field for 3D Point Cloud Interpolation, which implicitly integrates multi-frame information to handle nonlinear large motions for both indoor and outdoor scenarios. Furthermore, we construct a new multi-frame point cloud interpolation dataset called NL-Drive for large nonlinear motions in autonomous driving scenes to better demonstrate the superiority of our method. Ultimately, NeuralPCI achieves state-of-the-art performance on both DHB (Dynamic Human Bodies) and NL-Drive datasets. Beyond the interpolation task, our method can be naturally extended to point cloud extrapolation, morphing, and auto-labeling, which indicates its substantial potential in other domains. Codes are available at https://github.com/ispc-lab/NeuralPCI.

Event cameras have the ability to record continuous and detailed trajectories of objects with high temporal resolution, thereby providing intuitive motion cues for optical flow estimation. Nevertheless, most existing learning-based approaches for event optical flow estimation directly remould the paradigm of conventional images by representing the consecutive event stream as static frames, ignoring the inherent temporal continuity of event data. In this paper, we argue that temporal continuity is a vital element of event-based optical flow and propose a novel Temporal Motion Aggregation (TMA) approach to unlock its potential. Technically, TMA comprises three components: an event splitting strategy to incorporate intermediate motion information underlying the temporal context, a linear lookup strategy to align temporally fine-grained motion features and a novel motion pattern aggregation module to emphasize consistent patterns for motion feature enhancement. By incorporating temporally fine-grained motion information, TMA can derive better flow estimates than existing methods at early stages, which not only enables TMA to obtain more accurate final predictions, but also greatly reduces the demand for a number of refinements. Extensive experiments on DSEC-Flow and MVSEC datasets verify the effectiveness and superiority of our TMA. Remarkably, compared to E-RAFT, TMA achieves a 6% improvement in accuracy and a 40% reduction in inference time on DSEC-Flow. Code will be available at https://github.com/ispc-lab/TMA.

Neural Radiance Fields (NeRFs) have achieved great success in the past few years. However, most current methods still require intensive resources due to ray marching-based rendering. To construct urban-level radiance fields efficiently, we design Deformable Neural Mesh Primitive (DNMP), and propose to parameterize the entire scene with such primitives. The DNMP is a flexible and compact neural variant of classic mesh representation, which enjoys both the efficiency of rasterization-based rendering and the powerful neural representation capability for photo-realistic image synthesis. Specifically, a DNMP consists of a set of connected deformable mesh vertices with paired vertex features to parameterize the geometry and radiance information of a local area. To constrain the degree of freedom for optimization and lower the storage budgets, we enforce the shape of each primitive to be decoded from a relatively low-dimensional latent space. The rendering colors are decoded from the vertex features (interpolated with rasterization) by a view-dependent MLP. The DNMP provides a new paradigm for urban-level scene representation with appealing properties: (1) High-quality rendering. Our method achieves leading performance for novel view synthesis in urban scenarios. (2) Low computational costs. Our representation enables fast rendering (2.07 ms/1K pixels) and low peak memory usage (110 MB/1K pixels). We also present a lightweight version that can run 33× faster than vanilla NeRFs, which is comparable to the highly-optimized Instant-NGP. Project page: https://dnmp.github.io/.

Multi-agent collaborative perception (MCP) has recently attracted much attention. It includes three key processes: communication for sharing, collaboration for integration, and reconstruction for different downstream tasks. Existing methods pursue designing the collaboration process alone, ignoring their intrinsic interactions and resulting in suboptimal performance. In contrast, we aim to propose a Unified Collaborative perception framework named UMC, optimizing the communication, collaboration, and reconstruction processes with the Multi-resolution technique. The communication introduces a novel trainable multi-resolution and selective-region (MRSR) mechanism, achieving higher quality and lower bandwidth. Then, a graph-based collaboration is proposed, conducting on each resolution to adapt the MRSR. Finally, the reconstruction integrates the multi-resolution collaborative features for downstream tasks. Since the general metric can not reflect the performance enhancement brought by MCP systematically, we introduce a brand-new evaluation metric that evaluates the MCP from different perspectives. To verify our algorithm, we conducted experiments on the V2X-Sim and OPV2V datasets. Our quantitative and qualitative experiments prove that the proposed UMC outperforms the state-of-the-art collaborative perception approaches.

Accurately credit rating on Interbank assets is essential for a healthy financial environment and substantial economic development. But individual participants tend to provide manipulated information in order to attack the rating model to produce a higher score, which may conduct serious adverse effects on the economic system, such as the 2008 global financial crisis. To this end, in this paper, we propose a novel selective-aware graph neural network model (SA-GNN) for defense the Interbank credit rating attacks. In particular, we first simulate the rating information manipulating process by structural and feature poisoning attacks. Then we build a selective-aware defense graph neural model to adaptively prioritize the poisoning training data with Bernoulli distribution similarities. Finally, we optimize the model with weighed penalization on the objection function so that the model could differentiate the attackers. Extensive experiments on our collected real-world Interbank dataset, with over 20 thousand banks and their relations, demonstrate the superior performance of our proposed method in preventing credit rating attacks compared with the state-of-the-art baselines.

Medical insurance plays a vital role in modern society, yet organized healthcare fraud causes billions of dollars in annual losses, severely harming the sustainability of the social welfare system. Existing works mostly focus on detecting individual fraud entities or claims, ignoring hidden conspiracy patterns. Hence, they face severe challenges in tackling organized fraud. In this paper, we proposed RDPGL, a novel Risk Diffusion-based Parallel Graph Learning approach, to fighting against medical insurance criminal gangs. In particular, we first leverage a heterogeneous graph attention network to encode the local context from the beneficiary-provider graph. Then, we devise a community-aware risk diffusion model to infer the global context of organized fraud behaviors with the claim-claim relation graph. The local and global representations are parallel concatenated together and trained simultaneously in an end-to-end manner. Our approach is extensively evaluated on a real-world medical insurance dataset. The experimental results demonstrate the superiority of our proposed approach, which could detect more organized fraud claims with relatively high precision compared with state-of-the-art baselines.

Offline safe reinforcement learning (RL) algorithms promise to learn policies that satisfy safety constraints directly in offline datasets without interacting with the environment. This arrangement is particularly important in scenarios with high sampling costs and potential dangers, such as autonomous driving and robotics. However, the influence of safety constraints and out-of-distribution (OOD) actions have made it challenging for previous methods to achieve high reward returns while ensuring safety. In this work, we propose a Variational Optimization with Conservative Eestimation algorithm (VOCE) to solve the problem of optimizing safety policies in the offline dataset. Concretely, we reframe the problem of offline safe RL using probabilistic inference, which introduces variational distributions to make the optimization of policies more flexible. Subsequently, we utilize pessimistic estimation methods to estimate the Q-value of cost and reward, which mitigates the extrapolation errors induced by OOD actions. Finally, extensive experiments demonstrate that the VOCE algorithm achieves competitive performance across multiple experimental tasks, particularly outperforming state-of-the-art algorithms in terms of safety.

Point cloud registration is a fundamental problem in 3D computer vision. Previous learning-based methods for LiDAR point cloud registration can be categorized into two schemes: dense-to-dense matching methods and sparse-to-sparse matching methods. However, for large-scale outdoor LiDAR point clouds, solving dense point correspondences is time-consuming, whereas sparse keypoint matching easily suffers from keypoint detection error. In this paper, we propose SDMNet, a novel Sparse-to-Dense Matching Network for large-scale outdoor LiDAR point cloud registration. Specifically, SDMNet performs registration in two sequential stages: sparse matching stage and local-dense matching stage. In the sparse matching stage, we sample a set of sparse points from the source point cloud and then match them to the dense target point cloud using a spatial consistency enhanced soft matching network and a robust outlier rejection module. Furthermore, a novel neighborhood matching module is developed to incorporate local neighborhood consensus, significantly improving performance. The local-dense matching stage is followed for fine-grained performance, where dense correspondences are efficiently obtained by performing point matching in local spatial neighborhoods of high-confidence sparse correspondences. Extensive experiments on three large-scale outdoor LiDAR point cloud datasets demonstrate that the proposed SDMNet achieves state-of-the-art performance with high efficiency.

Point cloud registration is a fundamental problem in 3D computer vision. Outdoor LiDAR point clouds are typically large-scale and complexly distributed, which makes the registration challenging. In this paper, we propose an efficient hierarchical network named HRegNet for large-scale outdoor LiDAR point cloud registration. Instead of using all points in the point clouds, HRegNet performs registration on hierarchically extracted keypoints and descriptors. The overall framework combines the reliable features in deeper layer and the precise position information in shallower layers to achieve robust and precise registration. We present a correspondence network to generate correct and accurate keypoints correspondences. Moreover, bilateral consensus and neighborhood consensus are introduced for keypoints matching, and novel similarity features are designed to incorporate them into the correspondence network, which significantly improves the registration performance. In addition, we design a consistency propagation strategy to effectively incorporate spatial consistency into the registration pipeline. The whole network is also highly efficient since only a small number of keypoints are used for registration. Extensive experiments are conducted on three large-scale outdoor LiDAR point cloud datasets to demonstrate the high accuracy and efficiency of the proposed HRegNet. The source code of the proposed HRegNet is available at https://github.com/ispc-lab/HRegNet2.

In mobile interactive systems, there exist several components that can move and interact with each other. The current methods fail to give a comprehensive description of their properties and have inadequate modeling and analysis capacity for them. This paper concludes three system properties called system connectivity, interaction soundness and data validity, and focuses on their modeling and analysis based on Variable Petri Nets (VPNs), which have recently been proposed and are able to describe their dynamic interactions. A VPN- based model including component nets and interaction structure nets is constructed given a mobile interactive system. It depicts its structure and event-driven dynamics. Three properties and their related analysis methods of a VPN-based model are presented. An example is given to demonstrate the proposed concepts and methods. Note to Practitioners—Due to the mobility and frequent disconnections, the correctness of mobile interaction systems, such as mobile robot systems and mobile payment systems, are often difficult to analyze. This paper introduces three critical properties of systems, called system connectivity, interaction soundness and data validity, and presents a related modeling and analysis method, based on a kind of Petri net called VPN. For a given system, a model including component nets and interaction structure nets is constructed by using VPNs. The component net describes the internal process of each component, while the interaction structure net reflects the dynamic interaction between components. Based on this model, three properties are defined and analyzed. The case study of a practical mobile payment system shows the effectiveness of the proposed method.

Local differential privacy (LDP) is a prevalent measure of privacy protection as it provides rigorous privacy guarantees and has been widely studied for statistical analysis, especially in frequency estimation. As a representative LDP-enabled frequency estimation algorithm, Google'sRandomized Aggregation Privacy-Preserving Ordinal Response(RAPPOR) has been put into practice. However, it achieves sub-optimal utility due to the following limitations. Firstly, the adoption of the MD5 hash function inevitably results in the hash collision. Secondly, the application of the randomized response technique leads to randomness. To improve the practical effectiveness of RAPPOR and the utility of frequency-based services, we propose an LDP-enabled frequency estimation method called PK-RAPPOR, in which we devise an effective re-encoding hash function (RE-HF) incorporating prior knowledge (PK) about the rough frequency ranking of items. RE-HF divides items into several cohorts based on the PK and generates a unique hash value set for each item. Compared with the original RAPPOR, the hash collision can be eliminated for items from different cohorts, and the effect of randomness can be decreased by the overlapping of items from the same cohorts. We validate our proposed method with theoretical analysis and demonstrate its effectiveness with experiments on both synthetic and real-world datasets.

Gang fraud, the major and primary security issue in online lending services, can be efficiently solved by the data-driven paradigm that is recognized as a promising solution for online lending gang fraud prediction. However, it is challenging that such predictions need to detect evolving and increasingly impalpable fraud patterns based on low-quality data, i.e., very preliminary and coarse applicant information. The technical difficulty mainly stems from two factors: the extremedeficiency of information associationsandweakness of data labels. In this work, we mainly address the challenges by enhancing the utility of associations (i.e.,recovering missing associationsandmining underlying associations) on a knowledge graph. Specifically, we first propose an efficient method of Chinese address disambiguation to recover some critical associations that are broken by the ambiguity of applicant information, e.g., address related information. Then, to mine the implicit associations, we design a novel association representation method, calledAdaptive Connected Component Embedding Simplification Scheme(ACCESS), which can adaptively implement embedding for different connected components depending on their sizes. Finally, we adopt the graph clustering algorithms and devised predicting schemes based on the above enhanced associations to predict gang fraud in the case of weakness of data labels. Moreover, we propose a framework called RMCP by integrating the above techniques, which is consists of four steps:Recovering,Mining,Clustering, andPredicting, for efficiently predicting gang fraud. The good performance is validated by the experiments on a real-world dataset from a commercial lending company. Meanwhile, we provide a visual decision support system namedLongArmsover the RMCP framework.

Microservices and Kubernetes are widely used in the development and operations of cloud-native applications. By providing automated placement and scaling, Kubernetes has become the main tool for managing microservices. However, existing work and Kubernetes fail to consider the dynamic competition and availability of microservices as well as the problem of shared dependency libraries among multiple microservice instances. To this end, this article proposes an integer nonlinear microservice placement model for Kubernetes with the goal of cost minimization. Specifically, we calculate the number of instances based on microservice availability and construct a model in which the total resource demand of multiple microservice instances exceeds the appropriate proportion of node resources when dynamic resource competition exists and the size of the shared dependency library is less than the node storage capacity. Finally, this article solves the microservice placement model using an improved genetic algorithm. The experimental results demonstrate that higher throughput is obtained with the same costs and that the same throughput is obtained with lower costs.

The interaction process in microservice architectures is highly complex, making it very challenging to ensure correct behavioral interactions. The few related works focus only on the verification of interaction soundness in the case of a specific bufferk-value, without considering how to get the suitable bufferk-value. To solve the above problems, this article proposes a method to find the maximumk-value for microservice systems with bounded buffers, which can maximize the analysis of valuable asynchronous interaction paths while avoiding the waste of computer memory resources. Specific contributions include, first, giving the relationship between bufferk-value and asynchronous interaction paths, interaction soundness, and its proof; second, proposing an iterative detection based on the additional added paths algorithm and its correctness proof, which leads to the conclusion that finding the maximumk-value is a decidable problem; finally, validating the proposing methods on ten classical cases of microservice systems, and analyzing effectiveness and performance. The experimental results show that the method can effectively find the maximumk-value of bounded buffers compared with existing methods and thus ensure the correct behavioral interactions of microservice systems.

With the popularity of credit cards worldwide, timely and accurate fraud detection has become critically important to ensure the safety of their user accounts. Existing models generally utilize original features or manually aggregated features as their transactional representations, while they fail to reveal the hidden fraudulent behaviors. In this work, we propose a novel model to extract transactional behaviors of users and learn new transactional behavioral representations for credit card fraud detection. Considering the characteristics of transactional behaviors, two time-aware gates are designed in a recurrent neural net unit to learn long- and short-term transactional habits of users, respectively, and to capture behavioral changes of users caused by different time intervals between their consecutive transactions. A time-aware-attention module is proposed and employed to extract the behavioral information from their consecutive historical transactions with time intervals, which enables the proposed model to capture behavioral motive and periodicity inside their historical transactional behaviors. An interaction module is designed to learn more comprehensive and rational representations. To prove the effectiveness of the learned transactional behavioral representations, experiments are conducted on a large real-world transaction dataset and a public one. The results show that the learned representation can well distinguish fraudulent behaviors from legitimate ones, and the proposed method can improve the performance of credit card fraud detection in terms of various evaluation criteria over the state-of-the-art methods.

Graph neural networks (GNNs) have achieved effective performance in many graph-related tasks involving recommendation systems, social networks, and bioinformatics. Recent studies have proposed several graph pooling operators to obtain graph-level representations from node representations. Nevertheless, they usually adopt a single strategy to evaluate the importance of nodes, which may generate node rankings with weak robustness. Also, they cannot capture the different substructures of a graph since they shrink the graph layer by layer. To solve the above problems, this article proposes a Multistructure graph classification method with Attention mechanism and Convolutional neural network (CNN), called MAC. In particular, we propose a novel pooling operator, which adopts multiple strategies to evaluate the importance of nodes and updates node representations through an attention mechanism. Also, we design a hierarchical architecture for MAC to capture multiple different substructures of a graph. To further reduce the loss of graph information, we utilize 2-D CNN to generate a graph-level representation. Comparative experiments are performed on public benchmark datasets deriving from social systems, and the experimental results indicate that our method outperforms a range of state-of-the-art graph classification methods.

Open-set domain adaptation (OSDA) aims to achieve knowledge transfer in the presence of both domain shift and label shift, which assumes that there exist additional unknown target classes not presented in the source domain. To solve the OSDA problem, most existing methods introduce an additional unknown class to the source classifier and represent the unknown target instances as a whole. However, it is unreasonable to treat all unknown target instances as a group since these unknown instances typically consist of distinct categories and distributions. It is challenging to identify all unknown instances with only one additional class. In addition, most existing methods directly introduce marginal distribution alignment to alleviate distribution shift between the source and target domains, failing to learn discriminative class boundaries in the target domain since they ignore categorical discriminative information in the adaptation. To address these problems, in this article, we propose a novel prototype-based shared-dummy classifier (PSDC) model for the OSDA. Specifically, our PSDC introduces an auxiliary dummy classifier to calibrate the source classifier and simultaneously develops a weighted adaptation procedure to align class-wise prototypes for adaptation. We further design a pseudo-unknown learning algorithm to reduce the open-set risk. Extensive experiments on Office-31, Office-Home, and VisDA datasets show that the proposed PSDC can outperform existing methods and achieve the new state-of-the-art performance. The code will be made public.

In data-driven anti-fraud engineering for online payment services, the integration of proper function modules is an effective way to further improve detection performance by overcoming the inability of single-function methods to cope with complex and varied frauds. However, a qualified integration is really inaccessible under multiple demanding requirements, i.e., improving detection performance, ensuring decision explainability, and limiting processing latency and computing consumption. In this work, we propose a qualified integration system, named CAeSaR, that can simultaneously meet all of the above requirements. This satisfactory result is achieved by the cooperation of two innovative techniques. The first is a novel three-way taxonomy of function division, called TRTPT, according to the temporal positions of transactions relative to a reference fraudulent transaction. Based on TRTPT, CAeSaR can introduce three kinds of anti-fraud function modules which collaboratively cover all types of frauds theoretically. The second is an effective integration scheme, called TELSI. It generates the candidate decision strategies by combining the judgments of three function modules by only two simple logical connectives, which essentially ensures the decision explainability. Particularly, TELSI can assign the most effective decision strategy to the corresponding transaction adaptively by a devised stacking-based multi-classification. The advantages of CAeSaR are validated in practice over real-life data from a prestigious bank.

Formulating a cooperative autonomous vehicle group is challenging in an urban scene that has complex road networks and diverse disturbance. Existing methods of vehicle cluster cooperation in a vehicular ad-hoc network cannot be applied to autonomous vehicles because the latter have different requirements for a vehicle group structure and communication quality. Existing studies focus on autonomous vehicle group cooperation in closed and highway scenes only. Their outcomes cannot be directly applied to an urban scene because of its complex road conditions, incomplete cooperation properties, and lack of a vehicle group size control strategy. In this work, we formulate a cooperation model for autonomous vehicle groups in such scene. First, we analyze cooperation criteria based on the non-colliding aggregate motion of flocks and deduce the connectivity, coupling, timeliness, evolvability, and adaptivity of a vehicle group, based on which we propose a cooperation model. Next, we solve our model by using a modified distributed evolutionary multi-objective optimization method, prove its convergence, and analyze its computational complexity. Finally, we conduct simulations on synthetic and real roads to show its performance in terms of average connectivity, coupling, timeliness, evolvability, and adaptivity of vehicle groups.

Forecasting the future environment is an essential and fundamental capability of autonomous driving systems. In contrast to the widely studied video prediction, only a few literatures have explored LiDAR point cloud prediction. To achieve future point cloud generation, most existing methods are based on free-form 3D scene flow prediction. However, these simple scene flow prediction-based methods may cause distortions since the motions of 3D scenes can be seen as a combination of rigid (static background) and flow (dynamic foreground) motions. To address this issue, we propose a simple but effective rigid constrained point cloud prediction network named RPP-Net. The key component of the proposed method is the hybrid motion decoder, which generates motion masks and combines flow motions and rigid motions to produce hybrid motions without additional annotations and sensors. Besides, to reduce running time, we provide a hierarchical cell, which uses a feature encoder to extract deep features and an RPP-RNN module to capture temporal correlations across frames. To evaluate the effectiveness of our RPP-Net, we have conducted extensive experiments on both KITTI dataset and Argoverse dataset and the results show that our RPP-Net significantly outperforms existing methods and achieves a new state-of-the-art.

In networked-loans, guarantor enterprises have a legal duty to repay debt to the commercial bank when the guaranteed borrower enterprise defaults (fail to repay). During an economic recession, the risk of defaults may spread like wildfire -- the loan network structure could amplify both reach and impact; thus leading to a large-scale corporation defaults even systemic financial crises. The Central Bank urges advanced regulation technology to recognize and act on the contagion risk in order to avoid the ‘`gray rhino’'. Therefore, we present a novel approach to help the regulators quantify the systemic risk and provide stemming clues. In particular, we report a state-of-the-art graph neural network architecture (iConReg) for detecting and isolating of contagion risk in China's national-wide networked-loans. The overall accuracy of our model reaches over 91% of AUC, which considerably outperforms the compared benchmark methods. By isolating the top 1% of predicted high-risky nodes in the contagion chains, iConReg reports a significant shrink (averaged 25.8%) of loan default rates. Our presented approach opens up a new direction of using deep graph learning techniques to regulate the contagion risk of networked-loans, which enables the authorities to design more prompt prevention measures against systemic financial crises

Anti-money laundering (AML) is a classical data mining problem in finance applications. As well known, money laundering (ML) is critical to the effective operation of transnational and organized crime, which affects a country's economy, government, and social wellbeings. Financial services organizations facilitate the movement of money and have been enlisted by governments to assist with the detection and prevention of money laundering, which is a key tool in the fight to reduce crime and create sustainable economic development. In the application of AML, user identity and financial behavior data are widely used to detect laundering transactions. In recent years, an increasing number of money laundering activities have been conducted by organized criminal gangs while most existing works still treat the actions of each account as independent identity behavior without considering the group-level conspired interactions. Therefore, in this paper, we propose a group-aware deep graph learning-based approach for organized money-laundering detection. In particular, we design a community-centric encoder to represent the nodes and attributes in user transaction graphs and derive the adjacent gang behaviors. Then, we devise a scheme of local enhancement to accommodate nodes with similar transaction features, which are aggregated into gangs for downstream detection. Extensive experiments on the real-world dataset from one of the largest bank card alliances worldwide show that our proposed method outperforms state-of-the-art methods in both offline and online modes, showing the effectiveness of money laundering detection with group-aware deep graph learning.

Lightweight convolutional neural networks (CNNs) rely heavily on the design of lightweight convolutional modules (LCMs). For an LCM, lightweight design based on repetitive feature maps (LoR) is currently one of the most effective approaches. An LoR mainly involves an extraction of feature maps from convolutional layers (CE) and feature map regeneration through cheap operations (RO). However, existing LoR approaches carry out lightweight improvements only from the aspect of RO but ignore the problems of poor generalization, low stability, and high computation workload incurred in the CE part. To alleviate these problems, this article introduces the concept of key features from a CNN model interpretation perspective. Subsequently, it presents a novel LCM, namely CEModule, focusing on the CE part. CEModule increases the number of key features to maintain a high level of accuracy in classification. In the meantime, CEModule employs a group convolution strategy to reduce floating-point operations (FLOPs) incurred in the training process. Finally, this article brings forth a dynamic adaptation algorithm ( α -DAM) to enhance the generalization of CEModule-enabled lightweight CNN models, including the developed CENet in dealing with datasets of different scales. Compared with the state-of-the-art results, CEModule reduces FLOPs by up to 54% on CIFAR-10 while maintaining a similar level of accuracy in classification. On ImageNet, CENet increases accuracy by 1.2% following the same FLOPs and training strategies.

With the rapid development of manufacturing and cloud computing, more and more emerged cloud services provide a promising way to perform complex requirements efficiently. Workflow offers an effective way to assemble disparate services and interacts with them to construct business logic for user requests. Meanwhile, workflow engines are responsible for the control of workflow execution. However, existing engines usually support interaction between workflows and services or computing resources by tight binding approaches, which lack flexibility and scalability. Therefore, a flexible and decoupled architecture is essential to support automatic workflow management. To fill this gap, this paper proposes a novelservice-centered cloud-native workflow architecture- SCAFE. The introduction of the service layer in SCAFE decouples the upper business services and lower execution resources, facilitating the independent and joint management of three execution objects (request - service - execution instance) involved in the cloud workflow lifecycle. We present a 2-stage scheduling model for the new architecture to support customized service optimization and resource-aware task scheduling. In addition, a fault-tolerant mechanism is integrated into resolving task execution exceptions quickly. We have successfully implemented a prototype tool to verify its flexibility and crucial functions, thus advancing the field workflow engines in cloud-native environments.

The workload predictor has attracted attention as a key component of the proactive service operation management framework. However, the request and resource workloads of cloud applications are highly dynamic. Existing approaches decompose the original workload into trend, seasonal, and random components, establish models accordingly, and then combine all outputs to generate results. Indeed, the random component usually has significant heteroscedasticity and noise, having little or even a negative effect on model accuracy improvement. In our model, trend and seasonal components are seen as macro workload changes, and the micro workload changes are obtained by an adaptive sliding window algorithm. Therefore, we propose an ensembling model named FAST for Forecasting workloads with Adaptive Sliding window and Time locality integration. Notably, we propose an adaptive sliding window algorithm that considers trend correlation, time correlation, and random fluctuations of workload for online regression to achieve higher accuracy with lower overhead; and for the error-based integration strategy, we propose a time locality concept for local-predictor behavior and develop a multi-class regression algorithm for model integration. Finally, we conduct experiments on Google cluster trace datasets which show FAST has better accuracy than all other state-of-the-art models for dynamic workloads.

Accurately processing dynamic evolution events is extremely challenging for autonomous vehicle groups in an urban scene, which can be disturbed by manned vehicles, roadside obstacles, traffic lights, and pedestrians. Existing work focuses on a dynamic evolution method for such groups in a highway scene only. Its outcomes cannot be directly used to an urban scene due to different environmental factors, incomplete dynamic evolution events, and lack of simulation evaluation with real road networks. In this work, we present a dynamic evolution method for such groups in an urban scene. First, we analyze their dynamic evolution reasons. Then, we abstract five dynamic evolution events, i.e., joining, leaving, merging, splitting, and disappearing, and introduce a dynamic evolution method to process them. Finally, we deduce the evolvability that can reflect dynamic evolution states of a vehicle group. The simulation results in synthetic and real urban scenes show that the connectivity, coupling, timeliness, and evolvability of vehicle groups using the proposed dynamic evolution method are higher than those of using a dynamic evolution method for a highway scene.

The task of visual dialog requires a multimodal chatbot to answer sequential questions from humans about image content. Prior work performs the standard likelihood training for answer generation on the positive instances (involving correct answers). However, the likelihood objective often leads to frequent and dull outputs and fails to exploit the useful knowledge from negative instances (involving incorrect answers). In this paper, we propose a Unified Multimodal Model with UnLikelihood Training, named UniMM-UL, to tackle this problem. First, to improve visual dialog understanding and generation by multi-task learning, our model extends ViLBERT from only supporting answer discrimination to holding both answer discrimination and answer generation seamlessly by different attention masks. Specifically, in order to make the original discriminative model compatible with answer generation, we design novel generative attention masks to implement the autoregressive Masked Language Modeling (autoregressive MLM) task. And to attenuate the adverse effects of the likelihood objective, we exploit unlikelihood training on negative instances to make the model less likely to generate incorrect answers. Then, to utilize dense annotations, we adopt different fine-tuning methods for both generating and discriminating answers, rather than just for discriminating answers as in the prior work. Finally, on the VisDial dataset, our model achieves the best generative results (69.23 NDCG score). And our model also yields comparable discriminative results with the state-of-the-art in both single-model and ensemble settings (75.92 and 76.17 NDCG scores).

Vehicle groups that are composed of autonomous vehicles can increase the perception range of vehicles, and their dynamic evolution can provide guidance for the operation of autonomous vehicles. Most existing studies on vehicle group formation neither propose a standard vehicle group model, nor consider vehicle mobility and dynamic topology of vehicle groups. Instead, they focus on detecting dynamic evolution without predicting it. This work proposes a dynamic evolution method for autonomous vehicle groups. It first defines five vehicle states and their transitions. Then, it proposes an autonomous vehicle group formation method based on vehicle states and formulates an autonomous vehicle group model. Next, it uses meta vehicle group sequences to manage vehicle groups at different times. Finally, it gives detection and prediction methods of vehicle group dynamic evolution. Extensive simulation results show that the proposed method can be used to establish interconnection among autonomous vehicle nodes, detect dynamic evolution characteristics inside a vehicle group precisely, and predict dynamic evolution trends of vehicle groups effectively.

For the future open 5G Internet of Vehicles (IoV), due to the flexibility and load sharing, the popular network architecture of IoV proposed by many studies is the mobile-edge computing (MEC) framework combining with software-defined networking (SDN). However, under this architecture, moving vehicles and MEC devices are not like the cloud SDN with strong security protection. Thus, identity verification is an important security issue. We find that if the identity credentials of vehicles and infrastructures are obtained by adversaries (i.e., identity theft), the current cryptography-based authentication methods cannot cope with this problem. In this article, we propose a behavior-based verification method, named Crowd-Learning, by utilizing the idea of crowd in software-defined vehicular networks with a MEC framework. In Crowd-Learning, we design an incentive mechanism to stimulate some MEC infrastructures to provide accurate and appropriate amount of data for future correct behavior estimation. Without knowing the model of the dynamic environment, this incentive mechanism needs to apply reinforcement learning to let MEC infrastructures learn how to send data based on the current state. Our Crowd-Learning method verifies vehicles and reduces the verification latency by estimating the vehicle’s behavior in advance. Meanwhile, it verifies infrastructures during the process of reinforcement learning based on the idea of crowd intelligence. The fake infrastructures and anomalous vehicles expose themselves when learning. In experiments, we use the traffic simulation tool, called simulation of urban mobility (SUMO), to generate extensive vehicle traces and evaluate the performance of the Crowd-Learning verification method. The results show that the Crowd-Learning verification method can ensure high verification accuracy for vehicles and infrastructures with satisfying low verification latency.

How to guarantee the consistency between different systems has become a bone of contention in model-driven software development. Behavioral relations are a prominent technique/method to analyze the consistency degree of two workflow systems. There are four kinds of behavioral relations, i.e., strict order relation, exclusiveness relation, interleaving order relation, and inverse order relation. An activity in a module may be in different behavioral relations with activities in another module. We call the latter module aspecial modulein this article. However, the existing measuring methods for behavioral consistency do not consider special modules at all. Thus, if we conduct a consistency analysis of different systems by these methods, we may get an inaccurate consistency degree, or obtain no results. In order to solve this problem, we use workflow nets (WF-nets) to model workflow systems and re-explore their behavioral relations. Then, their behavioral relation matrices (BRMs) are constructed. Based on BRMs, some new measurement methods are proposed to calculate the behavioral consistency degree of two WF-net systems. Finally, a group of experiments is done to show the effectiveness of our methods.

Accurate estimation of network capacity is very important for Vehicular Infrastructure-based NETwork (VINET) in an urban scene that may involve greatly dynamic typology and complex driving conditions. The node mobility, network behavior, and network scale of a VINET are different from those of a wireless network, and, therefore, the existing capacity estimation methods of wireless networks cannot be used to estimate VINET capacity. In addition, most existing studies on VINET capacity only derive asymptotic descriptions when the number of nodes is large enough. In this work, a novel approach is proposed for the modeling and calculating VINET capacity. More specifically, we first analyze communication characteristics in a VINET, and introduce two transmission modes, i.e., a vehicle-based mode and a Road Side Unit (RSU)-based one. Then, we propose a probability-based transmission mode selecting strategy with which vehicle nodes can choose either transmission mode independently and such choice is probabilistic. Next, we analyze the characteristics of an RSU-based mode, divide a VINET into a number of communities according to the position and communication range of RSUs, and derive the capacity contributed by an RSU-based mode. Then, we calculate the capacity contributed by a vehicle-based mode based on fluid mechanics. Finally, the VINET capacity can be calculated. The proposed VINET capacity estimation approach is validated to be consistent with simulation results.

With the development of 5G technology, mobile edge computing (MEC) is introduced into the construction of internet of vehicles (IoV). However, the distributed denial of services (DDoS) attacks become a serious problem in IoV under MEC. Although numbers of studies have been done on DDoS detection in common wired or wireless networks, they cannot satisfy the high dynamic requirement and cannot cope with the complex and diverse DDoS attacks in IoV. Fortunately, the data traffic flows in IoV exist potential and predictable space-time regularities. By employing reinforcement learning, we propose a feature adaption reinforcement learning approach based on the space-time flow regularities in IoV for DDoS mitigation, named FAST. In FAST, we elaborately design a combinational action space, and a reward function based on Kalman filter method and historical data traffic flows, which can make FAST to recognize DDoS attacks more quickly and accurately. Then through combining Q-learning and DDQN, FAST can select features and disconnect DDoS attacks adaptively according to the changes of the environment. In experiments, we evaluate the performance of FAST based on Shenzhen taxicab dataset. We simulate and inject DDoS attacks into Shenzhen taxicabs through two DDoS simulation tools named ‘ddosflowgen’ and ‘hping3’. The experimental results show that FAST has a high quality in detecting multiple types of DDoS attacks compared with other detection methods.

Graph neural networks, which generalize deep learning to graph-structured data, have achieved significant improvements in numerous graph-related tasks. Petri nets (PNs), on the other hand, are mainly used for the modeling and analysis of various event-driven systems from the perspective of prior knowledge, mechanisms, and tasks. Compared with graph data, net data can simulate the dynamic behavioral features of systems and are more suitable for representing real-world problems. However, the problem of large-scale data analysis has been puzzling the PN field for decades, and thus, limited its universal applicability. In this article, a framework of net learning (NL) is proposed. NL contains the advantages of PN modeling and analysis with the advantages of graph learning computation. Then, two kinds of NL algorithms are designed for performance analysis of stochastic PNs, and more specifically, the hidden feature information of the PN is obtained by mapping net information to the low-dimensional feature space. Experiments demonstrate the effectiveness of the proposed model and algorithms on the performance analysis of stochastic PNs.

Recently, containers have become the primary deployment form for cloud applications. Predicting container workload accurately is critical to ensure the quality of service (QoS) and cost-efficiency of the applications and meet service level agreements (SLAs) with users. However, facing multiple challenges, including model unavailability due to insufficient data, model maladaptation due to dynamic workload changes, and model non-generalization due to changeable workload patterns in container workload prediction, existing methods have not yet provided a united and effective solution. To this end, we propose a novel integrated forecasting model named COIN that combines COmmon and INdividual changes in container workloads to ensure the availability, adaptivity, and generality of the prediction model based on transfer learning and online learning. Besides, we present a container similarity calculation algorithm for real cloud scenarios, which combines the static and dynamic information of containers and comprehensively depicts the similarity between containers. Through experiments based on two public datasets, the COIN model achieves a higher accuracy than existing state-of-the-art solutions, demonstrating the effectiveness and robustness of our proposed model, which provides a new solution to container workload prediction.

Mobile computing systems, service-based systems, and some other systems with mobile interacting components have recently received much attention. However, because of their characteristics, such as mobility and disconnection, it is difficult to model and analyze them by using a structure-fixed model. This work proposes a new Petri net model called variable Petri net (VPN) for modeling and analyzing these systems. The definition, firing rule, and related analysis technology of VPN are introduced in detail. In a VPN, the possible interaction interfaces are abstracted as a new kind of places called virtual places, and the occurrences of (dis)connections are described by new functions, which makes it appropriate to describe the component collaboration in systems and realize the scalability and pluggability of systems. Moreover, to overcome the shortcoming that markings cannot reflect the link capability of a system, VPNs add a constraint function along with a marking to represent a complete system configuration. Several examples are used to demonstrate the newly proposed model and method.

The properties of dynamic interactions in mobile-interactive systems are still difficult to analyze because of the complexity of systems. Thus, we have proposed a new Petri net called the variable petri net (VPN) recently, which specializes in describing dynamic interactions in systems. To make better use of VPN, this article focuses on the model checking method of VPN. It introduces the algorithm to transform a VPN to a Kripke structure that can describe both the system running states and the system connection states in VPN, and the method to transform a property to a temporal logic formula based on VPN and its Kripke structure. The Kripke structure can be optimized by considering the specific property about the system connection states and then be used to perform the targeted verification to the property by using a model checker. A practical example is given to demonstrate the proposed methods.

The existing Stochastic Petri Net (SPN) analysis methods are based on a series of steps, including generating the reachable graph, solving the state equation, etc. Unfortunately, these methods cannot perform performance analysis when the state equation has no unique solution. The end-to-end deep learning methods can build a mapping relationship from SPN to performance indicators, which avoids solving the state equation. However, there is a lack of benchmark datasets for SPN learning and training. This paper proposes an automatic generation method of SPN datasets, including SPN random generation, data labeling, data enhancement, and filtering. To relieve the local aggregation problem of random-based organization, a grid-based data organization method is proposed to ensure the diversity of the datasets. In the experimental section, the generated datasets are trained and tested on three types of neural networks. The results show that the generated benchmark datasets are useful, and the increase of dataset size will significantly improve the learning performance.

Existing work on the Internet-of-Vehicles (IoV) community mainly focuses on the detection of IoV community using static network detection and evolution methods of complex networks. These methods are prone to over-centralization, high computational complexity, and poor stability during the evolution of a community. In this work, we present an IoV community model and its evolution mechanism in an urban scene. More specifically, we first propose an IoV community detection model based on node similarity merging. Then, we use a network increment-based strategy to analyze node increment, edge increment, and weight increment. Finally, we give a dynamic evolution mechanism of an IoV community. Simulation-based experimental evaluation results show that the proposed mechanism achieves better real-time performance and accuracy than existing methods.

Fraud detection has attracted significant attention in financial institutions, especially utilizing some artificial intelligent methods to automatically detect fraudulent transactions. With the study and application of intelligent fraud detection technology, scholars found that the representation learning method can reveal more information about fraud patterns, which is also crucial for detection task. Therefore, in this work, we present a novel method for detecting fraud transactions by combining two modules learning hidden information at different levels of data in a unified framework. To address and explore the deep representation of features of transaction behaviors, we propose a two-level attention model to capture them by integrating two data embeddings at the data sample level and the feature level. In particular, the sample-level attention model captures the detailed information more centrally that is difficult to determine; the feature-level attention model extends the information of feature dependences. We further combine them to train a final fraud detection model. Extensive experiments are conducted using a data set provided by a financial company in China and several public financial data sets. The results confirm the effectiveness of our proposed method in detecting fraudulent transactions compared with other state-of-the-art methods.

Big data can help with providing valuable perceptions into business activities and disclosing the potential benefits. Advances in machine learning and deep learning technologies make it easier to achieve significant performance in a wide range of domains from city planning and marketing analysis to credit evaluation and identity theft detection. However, it still requires great efforts in selecting efficient learning algorithms and precise model parameters that are deemed confidential in the light of experience. Also worth noting is that there is a fundamental gap between impracticable business requirements and the available value of data reflected. The data holder or data service provider may not have a clear understanding of data interference. The solution to these two problems depends on the capability of predicting the data utility in advance, which raises a fundamental question: to what degree is the data utility predictable? In this work, we present a primary analytical framework for information-theoretic bounds of data utility and utilize the current state-of-the-art and representative algorithms to obtain the achievable lower bounds on a real-world data set. The gap between theoretical upper bounds and achievable lower bounds indicates that the achievable lower bounds can still be optimized for performance.

Location-Based Services (LBS), which answer users’ location-dependent queries to Points of Interest (POI), have become popular along with mobile devices’ widespread use. While benefiting from convenience, users may suffer privacy leak risk, due to publishing sensitive information, i.e., locations, to servers. Previous studies have proposed a number of methods for protecting LBS. Through providing provable privacy protection, Private Information Retrieval (PIR) becomes well-known for its high effectiveness in protecting privacy. However, the PIR-protected LBS must cautiously handle the consequent transmission/computation cost and service error brought by adopting PIR, as both of them impact user experience. Then an important question arises: Can the user experience of PIR-protected LBS be greatly improved? We answer it by presenting a novel interface from PIR to LBS, gaining benefits from road networks. The proposed interface is comprised of two road partition methods fully considering POI’s distribution along roads, and two response-calculating methods to satisfy users’ requirements on driving distance. Since Vehicular Location Based Services contain adequate road network knowledge, we accomplish our work on them. The experimental results on a real dataset validate that our interface can improve PIR-protected LBS greatly in user experience while protecting privacy.

Microservices are widely used for flexible software development. Recently, containers have become the preferred deployment technology for microservices because of fast start-up and low overhead. However, the container layer complicates task scheduling and auto-scaling in clouds. Existing algorithms do not adapt to the two-layer structure composed of virtual machines and containers, and they often ignore streaming workloads. To this end, this article proposes an Elastic Scheduling for Microservices (ESMS) that integrates task scheduling with auto-scaling. ESMS aims to minimize the cost of virtual machines while meeting deadline constraints. Specifically, we define the task scheduling problem of microservices as a cost optimization problem with deadline constraints and propose a statistics-based strategy to determine the configuration of containers under a streaming workload. Then, we propose an urgency-based workflow scheduling algorithm that assigns tasks and determines the type and quantity of instances for scale-up. Finally, we model the mapping of new containers to virtual machines as a variable-sized bin-packing problem and solve it to achieve integrated scaling of the virtual machines and containers. Via simulation-based experiments with well-known workflow applications, the ability of ESMS to improve the success ratio of meeting deadlines and reduce the cost is verified through comparison with existing algorithms.

The correctness of workflow systems is dependent on not only their control-flows but also their data-flows.WFD-nets(Workflow net with data) can model both control-flow and data-flow of workflow systems, and are used to analyze/verify their correctness. The existing methods of analyzing/verifying WFD-nets are based on their reachability graphs. However, the reachability-graph-based methods have two flaws: state space explosion and pseudo states, since the guard functions are not handled perfectly. Note that every transition of a WFD-net is labeled by a guard function, which is a Boolean expressionw.r.t.data and can determine the progress of control-flow. In order to solve these problems, we propose aguard-driven reachability graphthat can both alleviate the state space explosion problem and avoid pseudo states. We describe the related algorithms and develop a tool. A group of experiments illustrate the advantage and effectiveness of our approach, and an example of property loan shows its usefulness.

Whereas traditional scientific applications are computationally intensive, recent applications require more data-intensive analysis and visualization to extract knowledge from the explosive growth of scientific information and simulation data. As the computational power and size of compute clusters continue to increase, the I/O read rates and associated network for these data-intensive applications have been unable to keep pace. These applications suffer from long I/O latency due to the movement of “big data” from the network/parallel file system, which results in a serious performance bottleneck. To address this problem, we proposed a novel approach called “ODDS” to optimize data-locality access in scientific data analysis and visualization. ODDS leverages a distributed file system (DFS) to provide scalable data access for scientific analysis. Through exploiting the information of underlying data distribution in DFS, ODDS employs a novel data-locality scheduler to transform a compute-centric mapping into a data-centric one and enables each computational process to access the needed data from a local or nearby storage node. ODDS is suitable for parallel applications with dynamic process-to-data scheduling and for applications with static process-to-data assignment. To demonstrate the efficacy of our methods, we present and evaluate ODDS in the context of two state-of-the-art, scientific-analysis applications-mpiBLAST and ParaView-along with the Hadoop distributed file system (HDFS) across a wide variety of computing platform settings. In comparison to existing deployments using NFS, PVFS, or Lustre as the underlying storage systems, ODDS can greatly reduce the I/O cost and double overall performance.

Credit card fraud detection is an important study in the current era of mobile payment. Improving the performance of a fraud detection model and keeping its stability are very challenging because users' payment behaviors and criminals' fraud behaviors are often changing. In this article, we focus on obtaining deep feature representations of legal and fraud transactions from the aspect of the loss function of a deep neural network. Our purpose is to obtain better separability and discrimination of features so that it can improve the performance of our fraud detection model and keep its stability. We propose a new kind of loss function, full center loss (FCL), which considers both distances and angles among features and, thus, can comprehensively supervise the deep representation learning. We conduct lots of experiments on two big data sets of credit card transactions, one is private and another is public, to demonstrate the detection performance of our model by comparing FCL with other state-of-the-art loss functions. The results illustrate that FCL outperforms others. We also conduct experiments to show that FCL can ensure a more stable model than others.

AdaBoost is a boosting-based machine learning method under the assumption that the data in training and testing sets have the same distribution and input feature space. It increases the weights of those instances that are wrongly classified in a training process. However, the assumption does not hold in many real-world data sets. Therefore, AdaBoost is extended to transfer AdaBoost (TrAdaBoost) that can effectively transfer knowledge from one domain to another. TrAdaBoost decreases the weights of those instances that belong to the source domain but are wrongly classified in a training process. It is more suitable for the case that data are of different distribution. Can it be improved for some special transfer scenarios, e.g., the data distribution changes slightly over time? We find that the distribution of credit card transaction data can change with the changes in the transaction behaviors of users, but the changes are slow most of the time. These changes are yet important for detecting transaction fraud since they result in a so-called concept drift problem. In order to make TrAdaBoost more suitable for the abovementioned case, we, thus, propose an improved TrAdaBoost (ITrAdaBoost) in this article. It updates (i.e., increases or decreases) the weight of a wrongly classified instance in a source domain according to the distribution distance from the instance to a target domain, and the calculation of distance is based on the theory of reproducing kernel Hilbert space. We do a series of experiments over five data sets, and the results illustrate the advantage of ITrAdaBoost.

Since unmanned aerial vehicles (UAVs) can be controlled remotely in the absence of a unified means of identity authentication, they are quite vulnerable for illegal control by unidentifiable users. Only by tracing the identity of UAV itself, or providing management to pilots, current UAV identity authentication mechanism is far from achieving “single machine for single person.” With the development of artificial intelligence, it is possible to achieve automatic UAV identification. Therefore, this article proposes a behavior-based intelligent UAV identification and security supervision. Based on location tracking and flying data acquisition provided by the airborne black box, the UAV's behavioral data are collected on real time. Then, a reliable identification of UAVs is completed through the behavioral modeling, and a warning is issued in the potential illegal cases. It provides the government with intelligent control and disposal decision basis for flying UAVs.

Convolutional neural networks (CNNs) have demonstrated great competence in feature representation, and then, achieved a good performance to many classification tasks. Cross-entropy loss, together with softmax, is arguably one of the most commonly used loss functions in CNNs (that is generally called softmax loss). However, the softmax loss can result in a weakly discriminative feature representation since it focuses on the interclass separability rather than the intraclass compactness. This article proposes a pairwise Gaussian loss (PGL) for CNNs that can well address the intraclass compactness through significantly penalizing those similar sample pairs with a relatively large distance. At the same time, PGL can still ensure a good interclass separability. Experiments show that PGL can guarantee that CNNs obtain a better classification performance compared to not only the softmax loss but also others often used in CNNs. Our experiments also show that PGL has a stable convergence for the stochastic gradient descent optimization method in CNNs and a good generalization ability for different structures of CNNs.

Workflow nets and their extensions with data flows, e.g., workflow nets with data and colored workflow nets, can well model the business processes of workflow systems, and their soundness guarantees that systems have neither deadlock nor livelock. However, some design requirements related to the changes of data values within multiple business cases are not reflected by these models, and thus, some errors of the modeled systems cannot be detected by them. In this article, we propose workflow nets with tables (WFT-nets) to model workflow systems that are closely related to the database (i.e., different users, records, and attributes). We define the firing rule of WFT-nets via data refinements in order to avoid infinite states caused by unbounded data. We use first-order linear temporal logic to represent the design requirements of records and attributes' values and then define the soundness of WFT-nets. We construct an algorithm to verify soundness. A tool is developed, and a group of experiments are done to illustrate the effectiveness of our methods.

This paper focuses on the behavior similarity of workflow nets (WF-nets). The similarity of two WF-nets reflects their consistent degree in behaviors. It explores the behavioral relations of subsets of transitions based on the interleaving semantics, and more accurate relations are defined than the existing work. Therefore, a more accurate similarity of two WF-nets (in their behaviors) can be obtained than that in the existing work that usually do not consider the loop and complex correspondence. By refining the interleaving relation in a behavioral profile into six types, this paper proposes the notion of a relation profile based on behavioral profile. Based on the relation profile of a WF-net, behavioral relation matrix can be constructed. Additionally, we refine the complex correspondence and generate a group of behavioral relation submatrices from the behavioral relation matrix. By using them we present a new formula to measure the behavior similarity of two WF-nets. Finally, examples illustrate that our method can measure the similarity degree more accurately.

The e-commerce business process net (EBPN) is a novel formal model for describing an e-commerce system and its interactive parts, such as shoppers, merchants, and the third-party payment platforms. Vulnerability analysis has a great impact on the trustworthiness of EBPN, which is an issue stemming from data inconsistency problems. Data inconsistency problems affect the consistency of the EBPN transaction analysis. The underlying causes of inconsistent data are closely related to concurrent operations, such as control flow and data flow. However, most of the existing detection methods have difficulties characterizing the vulnerabilities and interactions of control and data flows. In this paper, we propose a new method based on the dynamic data slice (DDS) that considers both transaction consistency and data state consistency. First, by analyzing control flow characteristics of EBPN, we obtain the dynamic slice. This dynamic slice is based on all paths of the EBPN reachability graph. Second, we perform the data inconsistency analysis by considering both transaction consistency and data-state consistency. Based on these, we construct a DDS to characterize the behavioral logic and the data-dependence information. The DDS acquires the dynamic data firing sequence. Based on that sequence and a given data marking, we can construct the DDSs for several types of EBPNs. Constructing the DDS can be completed in polynomial time. The DDS is designed to characterize the behavioral logic and data-dependence information. Based on these, we design a method to judge the data constraints. This method satisfies the EBPN need for transaction consistency by considering both the control and data states. In addition, according to the data-dependence information, we can lock the vulnerable regions caused by abnormal trading data in the system. Finally, we give a method to compute the vulnerability level.

The class imbalance problem exists in many real-world applications such as fraud detection, medical diagnosis and spam filtering, and seriously influences the performance of learning algorithms. Randomly undersampling is a famous method to solve the problem. However, it cannot well extract the samples nearby the cross-edge of majority and minority classes due to its randomness, while these samples are very important for a classifier since they influence the classification performance. In this paper, we propose a novel Gaussian Mixture Undersampling (GMUS for short). GMUS mainly contains three steps. Firstly, a Gaussian Mixture Model (GMM) is applied to fit the majority samples. Secondly, considering the probability density function (PDF) of predicted minority samples on the well-fitted GMM, the maximum of PDF is selected as the cross-edge of two classes. Finally, we undersample the majority samples near the cross-edge. We do experiments on 16 public datasets and the results demonstrate that GMUS can sample more informative instances and thus improve the performance of classifiers compared with the state-of-the-art undersampling methods. We also apply GMUS to the credit card fraud detection and obtain a good performance.

Driving safety is a persistent concern for urban dwellers who spend hours driving on road in ordinary daily life. Traditional driving hazard detection solutions heavily rely on onboard sensors (e.g., front and rear radars, cameras) with limited sensing range. In this article, we propose a proactive hazard warning system, called APP, which aims to alert drivers when there are vehicles with dangerous behaviors nearby. To this end, APP incorporates several basic techniques (e.g, tensor decomposition, similarity comparison) to estimate behavioral data of a driver based on sparse sampled GPS trace at first. Then, with the estimated unlabelled data, potential dangerous behaviors of a particular vehicle are identified and recognized with a Gaussian Mixture Model (GMM) based approach. We have implemented and evaluated our system with a dataset collected for 30 days from over 13,676 taxicabs. Our method shows on average 81% accuracy in potential dangerous behavior recognition.

In human-computer interactive systems (HCISs), there are not only autonomous robots completely controlled by computers but also semiautonomous robots requiring human control. To avoid the errors in a procedure of interaction, a control model is needed. This paper proposes a systematic strategy with specific algorithms to construct an interactive-control-model based on Petri nets owing to their ability to describe concurrence and other system features. Instead of cumbersome iterations of deadlock detection in the existing studies, this paper introduces the concept of implicit constraints and the related implicit-conflict-marking-search algorithm to excavate them. In the algorithm, only the status of a single robot is needed to analyze the system instead of the status of all system components, which is an important innovation in this paper since this can well help one resolve the state explosion issue. Several examples are provided to show the feasibility of the proposed method. The proposed idea in this paper can be readily applied to practical HCISs.

Location recommendation method is an important application in a location-based social network. At present, it is a trend to integrate different recommendation methods since they have their own advantages in capturing different preferences of users and an integrated method can generally provide a better performance than every individual. However, the existing integration policies do not learn user preferences in their integration processes so that they cannot make full use of the advantage of each method. Therefore, we propose a novel integration method: learning-to-rank-based integration. In our method, a confidence coefficient is applied for each user in the integration process, and these coefficients can well optimize recommendation performance. A learning-to-rank-based algorithm is designed to train the confidence coefficients. A group of experiments are done on a real large-scale check-in data set, and the results demonstrate that our method outperforms the state-of-the-art ones.

Querying an incident (i.e., an occurrence of seemingly minor importance) from coarse-grained driving log (i.e., GPS trace) has been a daunting task. For example, “Which restaurant did I drive by at exactly 4 pm yesterday?” The question seems very simple but is nontrivial, because the question is semantics-driven while the actual log data are GPS coordinate-based. Especially, the practical GPS log is very sparse and inaccurate for high-speed mobile objects such as vehicles. This paper seeks to answer any fuzzy query over sparse vehicles GPS data. Our system, called iLogBook, achieves these two goals by leveraging tensor technique and latent semantic analysis to high-precision trajectory recovery and similarity matching. We have implemented and evaluated the iLogBook with the GPS data of over 13, 798 taxicabs collected in eight days in Shenzhen, China. Our results show about 97% accuracy in trajectory inference. Moreover, the system handles about 90% daily queries among 10 marked drivers.

As a typical latent factor model, Matrix Factorization (MF) has demonstrated its great effectiveness in recommender systems. Users and items are represented in a shared low-dimensional space so that the user preference can be modeled by linearly combining the item factor vector$V$using the user-specific coefficients$U$. From a generative model perspective,$U$and$V$are drawn from twoindependentGaussian distributions, which is not so faithful to the reality. Items are produced to maximally meet users’ requirements, which makes$U$and$V$strongly correlated. Meanwhile, the linear combination between$U$and$V$forces a bijection (one-to-one mapping), which thereby neglects the mutual correlation between the latent factors. In this paper, we address the upper drawbacks, and propose a new model, named Correlated Matrix Factorization (CMF). Technically, we apply Canonical Correlation Analysis (CCA) to map$U$and$V$into a new semantic space. Besides achieving the optimal fitting on the rating matrix, one component in each vector ($U$or$V$) is also tightly correlated with every single component in the other. We derive efficient inference and learning algorithms based on variational EM methods. The effectiveness of our proposed model is comprehensively verified on four public datasets. Experimental results show that our approach achieves competitive performance on both prediction accuracy and efficiency compared with the current state of the art.

As MapReduce is becoming ubiquitous in large-scale data analysis, many recent studies have shown that the performance of MapReduce could be improved by different job scheduling approaches, e.g., Fair Scheduler and Capacity Scheduler. However, most exiting MapReduce job schedulers focus on the scenario that MapReduce cluster is stable and pay little attention to the MapReduce cluster with dynamic resource availability. In fact, MapReduce cluster resources may fluctuate as there is a growing number of Hadoop clusters deployed on hybrid systems, e.g., infrastructure powered by mix of traditional and renewable energy, and cloud platforms hosting heterogeneous workloads. Thus, there is a growing need for providing predictable services to users who have strict requirements on job completion times in such dynamic environments. In this paper, we propose, RDS, a Resource and Deadline-aware Hadoop job Scheduler that takes future resource availability into consideration when minimizing job deadline misses. We formulate the job scheduling problem as an online optimization problem and solve it using an efficient receding horizon control algorithm. To aid the control, we design a self-learning model to estimate job completion times. We further extend the design of RDS scheduler to support flexible performance goals in various dynamic clusters. In particular, we use flexible deadline time bounds instead of the single fixed job completion deadline. We have implemented RDS in the open-source Hadoop implementation and performed evaluations with various benchmark workloads. Experimental results show that RDS substantially reduces the penalty of deadline misses by at least 36 and 10 percent compared with Fair Scheduler and Earliest Deadline First (EDF) scheduler, respectively. In a Hadoop cluster running partially on renewable energy, the experimental result shows the green power based resource prediction approach can further reduce the penalty of deadline misses by 16 percent compared to Auto-Regressive Integrated Moving Average (ARIMA) prediction approach.

In this paper, we present an I/O efficient distributed approximation framework to support approximations on arbitrary sub-datasets of a large dataset. Due to the prohibitive storage overhead of caching offline samples for each sub-dataset, existing offline sample-based systems provide high accuracy results for only a limited number of sub-datasets, such as the popular ones. On the other hand, current online sample-based approximation systems, which generate samples at runtime, do not take into account the uneven storage distribution of a sub-dataset. They work well for uniform distribution of a sub-dataset while suffer low I/O efficiency and poor estimation accuracy on unevenly distributed sub-datasets. To address the problem, we develop a distribution aware method called CLAP (cluster sampling based approximation). Our idea is to collect the occurrences of a sub-dataset at each logical partition of a dataset (storage distribution) in the distributed system, and make good use of such information to enable I/O efficient online sampling. There are three thrusts in CLAP. First, we develop a probabilistic map to reduce the exponential number of recorded sub-datasets to a linear one. Second, we apply the cluster sampling with unequal probability theory to implement a distribution-aware method for efficient online sampling for a single or multiple sub-datasets. Third, we enrich CLAP support with more complex approximations such as ratio and regression using bootstrap based estimation beyond the simple aggragation approxiamtions. Forth, we add an option in CLAP to allow users specifying a target error bound when submitting an approximation job. Fifth, we quantitatively derive the optimal sampling unit size in a distributed file system by associating it with approximation costs and accuracy. We have implemented CLAP into Hadoop as an example system and open sourced it on GitHub. Our comprehensive experimental results show that CLAP can achieve a speedup by up to 20× over the precise execution.

NAND Flash based Solid State Disks (SSDs) are gaining tremendous popularity in today's storage market due to their unique erase-before-write feature. The Flash Translation Layer (FTL) in the SSDs redirects the incoming writes to a free physical address and manages a logical to physical address mapping table. However, this induces significant performance degradation to the SSDs. One of the main reasons is that current cache management in FTLs is mainly optimized for the temporal or spatial locality. However, because of multiple levels of data buffers in the whole storage architecture, the locality of internal disk I/O is relatively low. What's more, the increasing capacity of SSD not only generates large mapping tables, but also imposes high pressure on the efficiency of page-level address mapping. To overcome this limitation, we propose Correlation-Aware Page-level FTL, a.k.a CPFTL, which exploits I/O correlations in the workloads. In CPFTL, we develop a correlation-aware mapping table based on the correlation in read operations. We then build a correlation prediction table to support fast mapping entry lookup in the correlation-aware mapping table. Finally, we split read and write caches and build a skew-aware dirty entry index to improve the cache hit ratio and reduce the garbage collection overhead. Our emulator and prototype are open-sourced at: https://github.com/janzhou/SSD-Emulator. The experimental results show that CPFTL can reduce the average response time by 63.4 percent for read dominant workloads and 32.9 percent for transaction workloads.

Modern datacenter schedulers apply a static policy to partition resources among different tasks. The amount of allocated resource won't get changed during a task's lifetime. However, we found that resource usage during a task's runtime demonstrates high dynamics and it only reaches full usage at few moments. Therefore, the static allocation policy doesn't exploit the dynamic nature of resource usage, leading to low system resource utilization. To address this hard problem, a recently proposed task-consolidation approach packs as many tasks as possible on the same node based on real-time resource demands. However, this approach may cause resource over-allocation and harm application performance. In this paper, we propose and develop ECS, an elastic container based scheduler that leverages resource usage variation within the task lifetime to exploit the potential utilization and parallelism. The key idea is to proactively select and shift tasks backward so that the inherent paralleled tasks can be identified without over-allocation. We formulate the scheduling scheme as an online optimization problem and solves it using a resource leveling algorithm. We have implemented ECS in Apache Yarn and performed evaluations with various MapReduce benchmarks in a cluster. Experimental results show that ECS can efficiently utilize resource and achieves up to 29% reduction on average job completion time while increasing CPU utilization by 25%, compared to stock Yarn.

Credit card is now popular in daily life. Meanwhile, credit card fraud events occur more frequently, which result in massive financial losses. There are a number of fraud detection methods, but they do not deeply mine features of customer's transaction behavior so that their detection effectiveness is not too desirable. This paper focuses on two aspects of feature mining. Firstly, the features of credit card transactions are expanded in time dimension to characterize the distinct payment habits of legal users and criminals. Secondly, Capsule Network (CapsNet) is adopted to further dig some deep features on the base of the expanded features, and then a fraud detection model is trained to identify if a transaction is legal or fraud. Through experiments on a real transaction dataset, we demonstrate that the time dimension extension can improve the performance of fraud detection, and then CapsNet is further illustrated to be more advantageous in fraud detection compared with other models.

Petri nets are used to detect the errors of data inconsistency in concurrent systems. The detection methods usually apply a breadth-first or depth-first search to the reachability graphs of Petri nets. However, they spend too much space and time due to the blind search strategy. In this paper, we propose a heuristic method that can save lots of space and time. First we define structural data inconsistency of Petri nets which is more likely to cause the errors of data inconsistency. Then, in order to detect the errors as fast as possible, a heuristic function is proposed to guide the construction of a reachability graph. Furthermore, we propose a detection algorithm and develop a tool. Our experiments illustrate the advantage and effectiveness of our method.

With the rapid development of electronic commerce, the number of transactions by credit cards are increasing rapidly. As online shopping becomes the most popular transaction mode, cases of transaction fraud are also increasing. In this paper, we propose a novel fraud detection method that composes of four stages. To enrich a cardholder's behavioral patterns, we first utilize the cardholders' historical transaction data to divide all cardholders into different groups such that the transaction behaviors of the members in the same group are similar. We thus propose a window-sliding strategy to aggregate the transactions in each group. Next, we extract a collection of specific behavioral patterns for each cardholder based on the aggregated transactions and the cardholder's historical transactions. Then we train a set of classifiers for each group on the base of all behavioral patterns. Finally, we use the classifier set to detect fraud online and if a new transaction is fraudulent, a feedback mechanism is taken in the detection process in order to solve the problem of concept drift. The results of our experiments show that our approach is better than others.

The Internet of Vehicles (IoV) is a subapplication of the Internet of Things in the automotive field. Large amounts of sensor data require to be transferred in real-time. Most of the routing protocols are specifically targeted to specific situations in IoV. But communication environment of IoV usually changes in the space-time dimension. Unfortunately, the traditional vehicular networks cannot select the optimal routing policy when facing the dynamic environment, due to the lack of abilities of sensing the environment and learning the best strategy. Sensing and learning constitute two key steps of the cognition procedure. Thus, in this paper, we present a software defined cognitive network for IoV (SDCIV), in which reinforcement learning and software defined network technology are considered for IoV to achieve cognitive capability. To the best of our knowledge, this paper is the first one that can give the optimal routing policy adaptively through sensing and learning from the environment of IoV. We perform experiments on a real vehicular dataset to validate the effectiveness and feasibility of the proposed algorithm. Results show that our algorithm achieves better performance than several typical protocols in IoV. We also show the feasibility and effectiveness of our proposed SDCIV.

The correctness of a time-critical system is closely related to the time of responding and performing every event. Our motivation example, alternating bit protocol , and application example, multi-track level crossing with sensors , both demonstrate that some nondeterministic behaviours can take place if the time associated with events is configured inappropriately or some concurrent events are controlled imperfectly in an overlapping period. These nondeterministic behaviours decrease the reliability and/or safety of a time-critical system. Therefore, it is valuable to formalise and check (non)determinacy. Time Petri Nets (TPN) in which the firing of every event is limited to a fix time interval are used to model time-critical systems in this article. We proposes a novel notion for TPN named time-soundness . It guarantees that the system always owns deterministic behaviours after any event is performed no matter when the event is performed. We utilise the notion of bisimulation to prove that the time-soundness can guarantee the behavioural determinacy. We propose an algorithm to check time-soundness, develop the related tool, and do experiments to show the usefulness and effectiveness of our notion and method.

With the popularization of online shopping, transaction fraud is growing seriously. Therefore, the study on fraud detection is interesting and significant. An important way of detecting fraud is to extract the behavior profiles (BPs) of users based on their historical transaction records, and then to verify if an incoming transaction is a fraud or not in view of their BPs. Markov chain models are popular to represent BPs of users, which is effective for those users whose transaction behaviors are stable relatively. However, with the development and popularization of online shopping, it is more convenient for users to consume via the Internet, which diversifies the transaction behaviors of users. Therefore, Markov chain models are unsuitable for the representation of these behaviors. In this paper, we propose logical graph of BP (LGBP) which is a total order-based model to represent the logical relation of attributes of transaction records. Based on LGBP and users' transaction records, we can compute a path-based transition probability from an attribute to another one. At the same time, we define an information entropy-based diversity coefficient in order to characterize the diversity of transaction behaviors of a user. In addition, we define a state transition probability matrix to capture temporal features of transactions of a user. Consequently, we can construct a BP for each user and then use it to verify if an incoming transaction is a fraud or not. Our experiments over a real data set illustrate that our method is better than three state-of-the-art oneoness.

Gene expression programming (GEP) is a data driven evolutionary technique that well suits for correlation mining. Parallel GEPs are proposed to speed up the evolution process using a cluster of computers or a computer with multiple CPU cores. However, the generation structure of chromosomes and the size of input data are two issues that tend to be neglected when speeding up GEP in evolution. To fill the research gap, this paper proposes three guiding principles to elaborate the computation nature of GEP in evolution based on an analysis of GEP schema theory. As a result, a novel data engineered GEP is developed which follows closely the generation structure of chromosomes in parallelization and considers the input data size in segmentation. Experimental results on two data sets with complementary features show that the data engineered GEP speeds up the evolution process significantly without loss of accuracy in data correlation mining. Based on the experimental tests, a computation model of the data engineered GEP is further developed to demonstrate its high scalability in dealing with potential big data using a large number of CPU cores.

Personalized trip recommendation has drawn much attention recently with the development of location-based services. How to utilize the data in the location-based social network to recommend a single Point of Interest (POI) or a sequence of POIs for users is an important question to answer. Recommending the latter is called trip recommendation that is a challenging study because of the diversity of trips and complexity of involved computation. This work proposes a maximal-marginal-relevance-based personalized trip recommendation method that considers both relevance and diversity of trips in trip planning. An ant-colony-optimization-based trip planning algorithm is developed to efficiently plan a trip. Finally, case studies and experiments illustrate the effectiveness of our method.

Document network is a kind of intriguing dataset which can provide both topical (textual content) and topological (relational link) information. A key point in modeling such datasets is to discover proper denominators beneath the text and link. Most previous work introduces the assumption that documents closely linked with each other share common latent topics. However, the heterophily (i.e., tendency to link to different others) of nodes is neglected, which is pervasive in social networks. In this paper, we simultaneously incorporate community detection and topic modeling in a unified framework, and appeal to Canonical Correlation Analysis (CCA) to capture the latent semantic correlations between the two heterogeneous factors, community and topic. Despite of the homophily (i.e., tendency to link to similar others) or heterophily, CCA can properly capture the inherent correlations which fit the dataset itself without any prior hypothesis. We also impose auxiliary word embeddings to improve the quality of topics. The effectiveness of our proposed model is comprehensively verified on three different types of datasets which are hyperlinked networks of web pages, social networks of friends, and coauthor networks of publications. Experimental results show that our approach achieves significant improvements compared with the current state of the art.

While Hadoop ecosystems become increasingly important for practitioners of large-scale data analysis, they also incur tremendous energy cost. This trend is driving up the need for designing energy-efficient Hadoop clusters in order to reduce the operational costs and the carbon emission associated with its energy consumption. However, despite extensive studies of the problem, existing approaches for energy efficiency have not fully considered the heterogeneity of both workload and machine hardware found in production environments. In this paper, we find that heterogeneity-oblivious task assignment approaches are detrimental to both performance and energy efficiency of Hadoop clusters. Our observation shows that even heterogeneity-aware techniques that aim to reduce the job completion time do not guarantee a reduction in energy consumption of heterogeneous machines. We propose a heterogeneity-aware task assignment approach, E-Ant, that aims to improve the overall energy consumption in a heterogeneous Hadoop cluster without sacrificing job performance. It adaptively schedules heterogeneous workloads on energy-efficient machines, without a priori knowledge of the workload properties. E-Ant employs an ant colony optimization approach that generates task assignment solutions based on the feedback of each task's energy consumption reported by Hadoop TaskTrackers in an agile way. Furthermore, we integrate DVFS technique with E-Ant to further improve the energy efficiency of heterogeneous Hadoop clusters. It relies on a DVFS controller to dynamically scale the CPU frequency of each slave machine in response to time-varying resource demands. Experimental results on a heterogeneous cluster with varying hardware capabilities show that E-Ant with DVFS improves the overall energy savings for a synthetic workload from Microsoft by 23 and 17 percent compared to Fair Scheduler and Tarazu, respectively.

Today enterprises have massive stream data that require to be processed in real time due to data explosion in recent years. Spark Streaming as an emerging system is developed to process real time stream data analytics by using micro-batch approach. The unified programming model of Spark Steaming leads to some unique benefits over other traditional streaming systems, such as fast recovery from failures, better load balancing and resource usage. It treats the continuous stream as a series of micro-batches of data and continuously process these micro-batch jobs. However, efficient scheduling of micro-batch jobs to achieve high throughput and low latency is very challenging due to the complex data dependency and dynamism inherent in streaming workloads. In this paper, we propose A-scheduler, an adaptive scheduling approach that dynamically schedules parallel micro-batch jobs in Spark Streaming and automatically adjusts scheduling parameters to improve performance and resource efficiency. Specifically, A-scheduler dynamically schedules multiple jobs concurrently using different policies based on their data dependencies and automatically adjusts the level of job parallelism and resource shares among jobs based on workload properties. Furthermore, we integrate dynamic batching technique with A-Scheduler to further improve the overall performance of the customized Spark Streaming system. It relies on an expert fuzzy control mechanism to dynamically adjust the length of each batch interval in response to time-varying streaming workload and system processing rate. We implemented A-scheduler and evaluated it with a real-time security event processing workload. Our experimental results show that A-scheduler with dynamic batching can reduce end-to-end latency by 38 percent and meanwhile improve workload throughput and energy efficiency by 23 and 15 percent, respectively, compared to the default Spark Streaming scheduler.

E-commerce business process nets (EBPNs) are a novel formal model for describing and validating e-commerce systems including interactive parties such as shopper, merchant, and third-party payment platform. Data errors and nondeterminacy of the data states during the trading process can be depicted with the help of EBPNs. However, the problem about how to analyze EBPNs remains largely open. To analyze their data-liveness, data-boundedness, and reachability, this paper presents two analysis methods. For EBPNs, reachability analysis is proposed based on a 3-D incidence matrix method. Additionally, reduction methods are proposed for a special EBPN. Finally, the validity and reliability of the proposed methods are illustrated via the examples of e-commerce systems.

Link similarity is widely applied in measuring the similarity between such objects as Web pages, scientific papers, and social networks. However, there are some deficiencies in the existing methods to measure it. For example, they cannot handle some semantic-similar contents. Their computation may not lead to accurate results in some cases. This paper presents a novel method to do so. It introduces the semantic similarity to calculate the similarity between two given objects, and overcomes the drawback caused by the fact that the existing methods ignore the semantic information of objects. It also gives a novel computation function to make the computing result of similarity more accurate.

GPUs are widely adopted in HPC and cloud computing platforms to accelerate general-purpose workloads. However, modern GPUs do not support flexible preemption, leading to performance and priority inversion problems in multi-tasking environments.

The soaring of social media services has greatly propelled the prevalence of document networks. Rather than a set of plain texts, documents are nodes in graphs. An observable link connects the documents at its two ends, thus it implicitly reflects the semantic association between the document pair. Previous work assumes that only similar documents tend to be connected, which neglects the rich connective patterns in the topological structure. In this paper, we introduce a latent correlation factor to categorize the links into several categories, and each category corresponds to a unique kind of association. By fitting the data, the relational information (e.g., homophily and heterophily) can be comprehensively captured. By resorting to Canonical Correlation Analysis (CCA), we maximize the correlation between all pairs of linked documents. We propose a pure generative model and derive efficient learning algorithms based on the variational EM methods. Experiments on three different datasets demonstrate that the proposed model is competitive and usually better than the state-of-the-art baselines on both topic modeling and link prediction.

Document network is a kind of intriguing dataset which provides both topical (texts) and topological (links) information. Most previous work assumes that documents closely linked with each other share common topics. However, the associations among documents are usually complex, which are not limited to the homophily (i.e., tendency to link to similar others). Actually, the heterophily (i.e., tendency to link to different others) is another pervasive phenomenon in social networks. In this paper, we introduce a new tool, called copula, to separately model the documents and links, so that different copula functions can be applied to capture different correlation patterns. In statistics, a copula is a powerful framework for explicitly modeling the dependence of random variables by separating the marginals and their correlations. Though widely used in Economics, copulas have not been paid enough attention to by researchers in machine learning field. Besides, to further capture the potential associations among the unconnected documents, we apply the tree-averaged copula instead of a single copula function. This improvement makes our model achieve better expressive power, and also more elegant in algebra. We derive efficient EM algorithms to estimate the model parameters, and evaluate the performance of our model on three different datasets. Experimental results show that our approach achieves significant improvements on both topic and link modeling compared with the current state of the art.

Summary Urban traffic condition usually serves as basic information for some intelligent urban applications, for example, intelligent transportation system. The traditional acquisition of such information is often costly because of the dependencies on infrastructures, such as cameras and loop detectors. Crowdsensing, as a new economic paradigm, can be utilized together with vehicular networks to efficiently gather vehicle‐sensed data for estimating the traffic condition. However, it has the problem of being lack of data uploading efficiency and data usage effectiveness. In this paper, we take into account the topology of the road net to deal with these problems. Specifically, we divide the road net intoroad sectionsandjunction areas. Based on this division, we introduce a two‐phased data collection and processing scheme named road topology‐based scheme. It leverages the correlations among adjacent roads. In a junction area, data collected by vehicles are first processed and integrated by a sponsor vehicle to locally calculate traffic condition. Both the selection of the sponsor and the calculation of road condition utilize the road correlation. The sponsor then uploads the local data to a server. By employing the inherent relations among roads, the server processes data and estimates traffic condition for the road sections without vehicular data in a global vision. We conduct experiments based on real vehicle trace data. The results indicate that our design can commendably handle the problems of efficiency and effectiveness in traffic condition evaluation using the vehicular crowdsensing data. Copyright © 2016 John Wiley & Sons, Ltd.

Fog computing, also called “clouds at the edge,” is an emerging paradigm allocating services near the devices to improve the quality of service (QoS). The explosive prevalence of Internet of Things, big data, and fog computing in the context of cloud computing makes it extremely challenging to explore both cloud and fog resource scheduling strategy so as to improve the efficiency of resources utilization, satisfy the users' QoS requirements, and maximize the profit of both resource providers and users. This paper proposes a resource allocation strategy for fog computing based on priced timed Petri nets (PTPNs), by which the user can choose the satisfying resources autonomously from a group of preallocated resources. Our strategy comprehensively considers the price cost and time cost to complete a task, as well as the credibility evaluation of both users and fog resources. We construct the PTPN models of tasks in fog computing in accordance with the features of fog resources. Algorithm that predicts task completion time is presented. Method of computing the credibility evaluation of fog resource is also proposed. In particular, we give the dynamic allocation algorithm of fog resources. Simulation results demonstrate that our proposed algorithms can achieve a higher efficiency than static allocation strategies in terms of task completion time and price.

While MapReduce is inherently designed for batch and high throughput processing workloads, there is an increasing demand for non-batch processes on big data, e.g., interactive jobs, real-time queries, and stream computations. Emerging Apache Spark fills in this gap, which can run on an established Hadoop cluster and take advantages of existing HDFS. As a result, the deployment model of Spark-on-YARN is widely applied by many industry leaders. However, we identify three key challenges to deploy Spark on YARN, inflexible reservation-based resource management, inter-task dependency blind scheduling, and the locality interference between Spark and MapReduce applications. The three challenges cause inefficient resource utilization and significant performance deterioration. We propose and develop a cross-platform resource scheduling middleware, iKayak, which aims to improve the resource utilization and application performance in multi-tenant Spark-on-YARN clusters. iKayak relies on three key mechanisms: reservation-aware executor placement to avoid long waiting for resource reservation, dependency-aware resource adjustment to exploit under-utilized resource occupied by reduce tasks, and cross-platform locality-aware task assignment to coordinate locality competition between Spark and MapReduce applications. We implement iKayak in YARN. Experimental results on a testbed show that iKayak can achieve 50 percent performance improvement for Spark applications and 19 percent performance improvement for MapReduce applications, compared to two popular Spark-on-YARN deployment models, i.e., YARN-client model and YARN-cluster model.

Petri nets are widely used to model and analyse concurrent systems. There exist two distinct classes of Petri nets that focus on different features of concurrent systems. The first one features multiple parallel processes sharing a group of common resources but not interacting/collaborating with each other. The second one allows multiple parallel processes to interact/collaborate with each other via message exchange but does not share any common resources. However, in many distributed environments, multiple processes both interact/collaborate with each other and share some common resources. To model and analyse such systems, this article defines a new class of Petri nets called Parallel Process Nets (P 2 Ns) that may be viewed as a generalization of the two mentioned above. We propose collaborativeness and close collaborativeness for P 2 Ns. The former guarantees that a modelled system is both deadlock-free and livelock-free, and the latter guarantees that it is deadlock-free, livelock-free, and starvation-free. These concepts and ideas are illustrated through some classical examples such as Producer-Consumer Problem and Dinning Philosophers Problem. Algorithms are developed to decide them. At last, P 2 Ns are applied to the modelling and analysis of two real systems: hospital information system and elevator scheduling system.

The errors of data inconsistency occur in a concurrent system when some concurrent operations are conducted improperly. The model-checking technique is widely used to detect them based on the state transition graph. However, the state space explosion problem is the biggest obstacle for this technique, since the state transition graph is based on the interleaving semantics that can result in a rapid increase of the graph scale. In addition, data inconsistency is closely related with concurrent operations, but the state transition graph hardly characterizes concurrency due to its interleaving semantics. The unfolding technique of Petri nets can both alleviate the state explosion and characterize concurrency because it is based on the concurrent semantics. In this paper, we define Petri net with data to model concurrent systems with three kinds of data operations: read , write, and delete, and then formalize data inconsistency. We propose an unfolding method to produce a finite complete prefix (FCP) for each PD-net. Then, a matrix that represents all concurrency relations of transitions is constructed in view of FCP. Furthermore, the error of data inconsistency can be detected via this matrix. The related algorithms and the developed tool are introduced, and experiments illustrate their effectiveness and advantages. An example of industrial information system shows the usefulness of our study.

Recently, a number of technologies have been developed to promote vehicular networks. When vehicles are associated with the heterogeneous base stations (e.g., macrocells, picocells, and femtocells), one of the most important problems is to make load balancing among these base stations. Different from common mobile networks, data traffic in vehicular networks can be observed having regularities in the spatial-temporal dimension due to the periodicity of urban traffic flow. By taking advantage of this feature, we propose an online reinforcement learning approach, called ORLA. It is a distributed user association algorithm for network load balancing in vehicular networks. Based on the historical association experiences, ORLA can obtain a good association solution through learning from the dynamic vehicular environment continually. In the long run, the real-time feedback and the regular traffic association patterns both help ORLA cope with the dynamics of network well. In experiments, we use QiangSheng taxi movement to evaluate the performance of ORLA. Our experiments verify that ORLA has higher quality load balancing compared with other popular association methods.

Internet of Vehicles (IoV) has gained considerable attention from industry and academia due to the development of communication technology and smart city. However, a proprietary and closed way of operating hardware in network equipment slows down the progress of new service deployment and extension in IoV. Moreover, the tightly coupled control and data planes in traditional networks significantly increase the complexity and cost of network management. By proposing a novel architecture, i.e., software-defined IoV (SDIV), we adopt a software-defined network (SDN) architecture to address these problems by leveraging its separation of the control plane from the data one and a uniform way to configure heterogeneous switches. However, IoV characteristics introduce some great challenges in rule installation due to the limited size of flow tables at OpenFlow-enabled switches that are the main SDN component. It is necessary to build compact flow tables for IoV scalability. Accordingly, we develop a novel rule installation mechanism to reduce the number of rules for real-time query services in SDIV. We separate the wired data plane from the wireless one and use multicast addresses in the latter. We introduce a destination-driven model in the wired data plane to reduce the number of rules at switches. Experiments with a real data trace show that the developed approach significantly reduces the number of rules without degrading the performance of data transmissions for real-time query services in IoV.

Datacenter-scale clusters are evolving toward heterogeneous hardware architectures due to continuous server replacement. Meanwhile, datacenters are commonly shared by many users for quite different uses. It often exhibits significant performance heterogeneity due to multi-tenant interferences. The deployment of MapReduce on such heterogeneous clusters presents significant challenges in achieving good application performance compared to in-house dedicated clusters. As most MapReduce implementations are originally designed for homogeneous environments, heterogeneity can cause significant performance deterioration in job execution despite existing optimizations on task scheduling and load balancing. In this paper, we observe that the homogeneous configuration of tasks on heterogeneous nodes can be an important source of load imbalance and thus cause poor performance. Tasks should be customized with different configurations to match the capabilities of heterogeneous nodes. To this end, we propose a self-adaptive task tuning approach, Ant, that automatically searches the optimal configurations for individual tasks running on different nodes. In a heterogeneous cluster, Ant first divides nodes into a number of homogeneous subclusters based on their hardware configurations. It then treats each subcluster as a homogeneous cluster and independently applies the self-tuning algorithm to them. Ant finally configures tasks with randomly selected configurations and gradually improves tasks configurations by reproducing the configurations from best performing tasks and discarding poor performing configurations. To accelerate task tuning and avoid trapping in local optimum, Ant uses genetic algorithm during adaptive task configuration. Experimental results on a heterogeneous physical cluster with varying hardware capabilities show that Ant improves the average job completion time by 31, 20, and 14 percent compared to stock Hadoop (Stock), customized Hadoop with industry recommendations (Heuristic), and a profilingbased configuration approach (Starfish), respectively. Furthermore, we extend Ant to virtual MapReduce clusters in a multi-tenant private cloud. Specifically, Ant characterizes a virtual node based on two measured performance statistics: I/O rate and CPU steal time. It uses k-means clustering algorithm to classify virtual nodes into configuration groups based on the measured dynamic interference. Experimental results on virtual clusters with varying interferences show that Ant improves the average job completion time by 20, 15, and 11 percent compared to Stock, Heuristic and Starfish, respectively.

Load imbalance is a major source of overhead in parallel programs such as MapReduce. Due to the uneven distribution of input data, tasks with more data become stragglers and delay the overall job completion. Running Hadoop in a private cloud opens up opportunities for expediting stragglers with more resources but also introduces problems that often outweigh the performance gain: (1) performance interference from co-running jobs may create new stragglers; (2) there exists a semantic gap between the Hadoop task management and resource pool-based virtual cluster management preventing tasks from using resources efficiently. In this paper, we strive to make Hadoop more resilient to data skew and more efficient in cloud environments. We presentFlexSlot, a user-transparent task slot management scheme that automatically identifies map stragglers and resizes their slots accordingly to accelerate task execution. FlexSlot adaptively changes the number of slots on each virtual node to balance the resource usage so that the pool of resources can be efficiently utilized. FlexSlot further improves mitigation of data skew with an adaptive speculative execution strategy. Experimental results show that FlexSlot effectively reduces job completion time up to$47.2$percent compared to stock Hadoop and two recently proposed skew mitigation and speculative execution approaches.

This paper proposes a fully expanded tree (FET) approach for one-place-unbounded Petri nets. The FET of a one-place-unbounded Petri net consists of all and only reachable markings from its initial marking. Its applications to liveness and deadlock analysis for such Petri nets are developed. The proposed method has a larger application scope than all the existing methods for them. Several examples are provided to show its superiority over the state-of-the-art methods.

Effectiveness is the most important factor considered in the ranking models yielded by algorithms of learning to rank (LTR). Most of the related ranking models only focus on improving the average effectiveness but ignore robustness. When a ranking model ignores robustness, the effectiveness for many queries is possibly very poor although the average effectiveness for all queries is relatively high. Therefore, Wang et al. first consider robustness in their ranking models. However, the robustness formula defined by Wang et al. cannot characterize those queries whose effectiveness are hurt seriously in comparison with the baseline model. In order to overcome this shortcoming, we propose a novel formula of characterizing robustness based on portfolio theory, and construct a multiobjective optimization model of the robust LTR in which the formula is used. Based on this model, we propose an approach of risk-sensitive and robust LTR, named as R2Rank, which is based on the framework of archived multiobjective simulated annealing algorithm and the idea of preference ranking organization method for enrichment evaluation. The experimental results show that the ranking models produced by our proposed R2Rank approach are better in both effectiveness and robustness than those produced by three state-of-the-art LTR approaches.

Liveness of Petri nets means all activities in a modeled system can potentially take place and thus implies deadlock freedom. The research on liveness analysis approaches is inadequate. This paper proposes a liveness judgment reachability graph (LJRG) approach to analyze the liveness of ω-independent Petri nets. Such nets can be loosely explained as a class of unbounded Petri nets in which the changes of tokens in unbounded places are not related to each other. This paper proposes several algorithms to transform a new modified reachability tree to a new modified reachability graph and then transform it to an LJRG. It then develops the application of LJRG into the liveness analysis of ω-independent unbounded Petri nets. The proposed method provides a new theoretical method and important tool for the liveness analysis of unbounded Petri nets. It is illustrated via some examples.

It is a valuable study for Location-based Social Network (LBSN) make a more accurate Points-of-Interest (POI) recommendation since that can improve users' experiences. There have been many methods of POIs recommendation that consider context, personal preference pattern, and/or matrix factorization. However, the continuous contexts have not been thoroughly considered in these methods. This paper first proposes a locations splitting method which can handle both continuous and discrete contexts. Moreover, we present a Context-aware Probabilistic Matrix Factorization method (CPMF) that factorizes a frequency matrix of contexts and locations to obtain the user-location checkin probabilities. We design a Personal Preference Confidence (PPC) to extract a set of reliable POIs with confidence values for every user. Finally, we propose a hybrid recommender which fuses CPMF with PPC to recommend top-n POIs. Experiments on a large-scale real-world checkins dataset demonstrate that our recommendation method obtains a well performance and effect.

The general Petri nets are not suitable to model the data operations of concurrent read and coverable write. Therefore, Petri net with data operations (PN-DO) is defined, which extends contextual nets with write arcs and some other components. Its execution semantics are defined, and a new method is proposed to construct its reachability graph that is of a smaller scale than traditional reachability graph. Based on this kind of reachability graph, we propose a method to check the errors of inconsistent data and missing data. Meanwhile, case studies are given to illustrate the effectiveness of our methods.

In this paper, we aim to model the formation of interest-driven data dissemination in online social networks (OSNs). We focus on a usual type of interest-driven social sessions in OSNs, called Social-InterestCast, under which a user will autonomously determine whether to view the content from his followees depending on his interest. To figure out the formation mechanism of such a Social-InterestCast, we propose a four-layered system model, consisting of physical layer, social layer, content layer, and session layer to model this interestdriven sessions. To the best of our knowledge, this is the first work to model data dissemination in OSNs with the interestdriven characteristics.

Traditional network anomaly detection involves developing models that rely on packet inspection. However, increasing network speeds and use of encrypted protocols make per-packet inspection unsuited for today’s networks. One method of overcoming this obstacle is aggregating packet header information and performing flow-based analysis where data flow patterns are examined rather than deep packet inspection. Many existing approaches are special purpose limited to detecting specific behavior. Also, the data reduction inherent in identifying anomalous flows hinders alert correlation. In this article, we propose and develop a dynamic anomaly detection approach for augmented network flows. We sketch network state during flow creation, enabling general-purpose threat detection. We describe an efficient flow augmentation approach based on the count-min sketch that provides per-flow-, per-node-, and per-network-level statistics parallel to flow record generation. We design and develop a support vector machine-based adaptive anomaly detection and correlation mechanism, which is capable of aggregating alerts without a priori alert classification and evolving models online. We further develop a lightweight evolving alert aggregation method and combine it with a confidence forwarding mechanism identifying a small percentage predictions for additional processing. We show effectiveness of our methods on both enterprise and backbone traces. Experimental results demonstrate its ability to maintain high accuracy without the need for offline training.

Recently, online shopping integrating third-party payment platforms (TPPs) introduces new security challenges due to complex interactions between Application Programming Interfaces (APIs) of Merchants and TPPs. Malicious clients may exploit security vulnerabilities by calling APIs in an arbitrary order or playing various roles. To deal with the security issue in the early stages of system development, this paper presents a formal method for modeling and verification of online shopping business processes with malicious behavior patterns considered based on Petri nets. We propose a formal model called E-commerce Business Process Net to model a normal online shopping business process that represent intended functions, and malicious behavior patterns representing a potential attack that violates the security goals at the requirement analysis phase. Then, we synthesize the normal business process and malicious behavior patterns by an incremental modeling method. According to the synthetic model, we analyze whether an online shopping business process is resistant to the known malicious behavior patterns. As a result, our approach can make the software design provably secured from the malicious attacks at process design time and, thus, reduces the difficulty and cost of modification for imperfect systems at the release phase. We demonstrate our approach through a case study.

While major Cloud service operators have taken various initiatives to operate their datacenters with renewable energy partially or completely, it is challenging to effectively utilize the renewable energy since its generation depends on dynamic natural conditions. In this paper, we propose and develop an elastic power-aware resource provisioning approach (ePower) for heterogeneous workloads in self-sustainable datacenters that completely rely on renewable energy. We aim to maximize the system goodput and control the system power consumption with respect to green power supply. ePower takes challenges and advantages of dynamic power supply, heterogeneous workload characteristics and QoS requirements, and automatically optimizes elastic resource allocations to workloads. The core of ePower design is a novel power-aware simulated annealing algorithm with fuzzy performance modeling for the efficient search of an optimal resource allocation. We have implemented ePower in a university cloud testbed hosting Gridmix2 and RUBiS benchmark applications. We utilize real weather data traces to simulate the green power generation and supply in the experiments. Experimental results demonstrate ePower can achieve near-to-optimal system performance while being resilient to dynamic power availability. It outperforms a representative resource provisioning approach for heterogeneous workloads by at least 24% in improving system goodput and 35 percent in reducing QoS violations.

MapReduce has become a major computing model for data intensive applications. Hadoop, an open source implementation of MapReduce, has been adopted by an increasingly growing user community. Cloud computing service providers such as Amazon EC2 Cloud offer the opportunities for Hadoop users to lease a certain amount of resources and pay for their use. However, a key challenge is that cloud service providers do not have a resource provisioning mechanism to satisfy user jobs with deadline requirements. Currently, it is solely the user's responsibility to estimate the required amount of resources for running a job in the cloud. This paper presents a Hadoop job performance model that accurately estimates job completion time and further provisions the required amount of resources for a job to be completed within a deadline. The proposed model builds on historical job execution records and employs Locally Weighted Linear Regression (LWLR) technique to estimate the execution time of a job. Furthermore, it employs Lagrange Multipliers technique for resource provisioning to satisfy jobs with deadline requirements. The proposed model is initially evaluated on an in-house Hadoop cluster and subsequently evaluated in the Amazon EC2 Cloud. Experimental results show that the accuracy of the proposed model in job execution estimation is in the range of 94.97 and 95.51 percent, and jobs are completed within the required deadlines following on the resource provisioning scheme of the proposed model.

In a datacenter, complex and time-varying interactions between various tiers and services of web applications, and the contention of shared resources among co-located virtual machines have significant impact on the user perceived performance and power consumption of the underlying system. We propose and develop APPLEware, an autonomic middleware for joint performance and power control of co-located web applications in virtualized datacenters. It features a distributed control structure that provides predictable performance and energy efficiency for large complex systems. It applies machine learning based self-adaptive modeling to capture the complex and time-varying relationship between the application performance and allocation of resources to various application components, in the face of highly dynamic and bursty workloads. The distributed controllers coordinate with each other and allocate resources to meet the service level agreements of applications in an agile and energy-efficient manner. Experimental results based on a testbed implementation with benchmark applications and large scale simulations demonstrate APPLEware's effectiveness, energy efficiency and scalability.

By arranging multiple existing web services into workflows to create value-added services, automatic web service composition has received much attention in service-oriented computing. A large number of methods have been proposed for it although most of them are merely based on the matching of input-output parameters of services. Besides these parameters, some other elements can affect the execution of services and their composition, such as the preconditions and service execution results. In particular, the execution effects of some services are often uncertain because of the complex and dynamically changing application environments in the real world, and this can cause the emergence of nondeterministic choices in the workflows of composite services. However, the previous methods for automatic service composition mainly rely on sequential structures, which make them difficult to take into account uncertain effects during service composition. In this paper, Graphplan is employed and extended to tackle this problem. In order to model services with uncertain effects, we first extend the original form of Graphplan. Then, we propose a novel approach that can introduce branch structures into composite solutions to cope with such uncertainty in the service composition process. Extensive experiments are performed to evaluate and analyze the proposed methodology.

The number of web services has grown drastically. Then how to manage them efficiently in a service repository is an important issue to address. Given a special field, there often exists an efficient data structure for a class of objects, e.g., the Google' Bigtable is very suitable for webpages' storage and management. Based on the theory of the equivalence relations and quotient sets, this work proposes a multilevel index model for large-scale service repositories, which can be used to reduce the execution time of service discovery and composition. Its novel use of keys as inspired by the key in relational database can effectively remove the redundancy of the commonly-used inverted index. Its four function-based operations are for the first time proposed to manage and maintain services in a repository. The experiments validate that the proposed model is more efficient than the existing structures, i.e., sequential and inverted index ones.

Due to the openness and dynamic characteristics of the network environment, the failure of component services in a composite Web service (CWS) can occur. CWS is a conglomeration of existing Web services (WSs) interacting together to offer a new value-added service. Hence, it is necessary to perform some transactions to ensure its correct and reliable execution. However, their introduction complicates its quality of service (QoS) analysis. How to analyze and evaluate its QoS values based on them becomes an important research issue. To address the problem, this paper introduces transactional properties of a single WS and CWS and conducts the performance analysis of basic workflow patterns such as sequential, parallel, selectable, and loop ones. Then, it develops an algorithm to compute the execution time of CWS and shows an example to illustrate the effectiveness of the proposed algorithm.

This correspondence paper focuses on classification and recognition of different objects in a satellite image. First, for every object we need to compute its fingerprint as its unique recognition. We improve the traditional elastic grid technique. Every object is partitioned into a set of grids. For each grid, we use its texture feature, a five-tuple features generated by gray level co-occurrence matrix, rather than its center value, an average value of grays, to characterize it. We utilize the feature-standardizing method to handle this five-tuple features and then generate the fingerprint of each grid. An ordered sequence of fingerprints of all grids of an object is viewed as the fingerprint of this object. Furthermore, on the basis of the fingerprints of objects, we use the Lebesgue measure to compute their dissimilarities, and thus these objects are classified. In this paper, we develop the related algorithms. The experimental results show that classification and recognition generated by our method is more accurate than others, which implies that the method of computing dissimilarity is better.

An exploratory search may be driven by a user's curiosity or desire for specific information. When users investigate unfamiliar fields, they may want to learn more about a particular subject area to increase their knowledge rather than solve a specific problem. This work proposes a topic-oriented exploratory search method that provides browse guidance to users. It allows them to discover new associations and knowledge, and helps them find their interested information and knowledge. Since an exploratory search needs to judge the ability to discover new knowledge, the existing commonly used metrics fail to capture it. This paper thus defines a new set of criteria containing clarity, relevance, novelty, and diversity to analyze the effectiveness of an exploratory search. Experiments are designed to compare results from the proposed method and Google's “search related to ....” The results show that the proposed one is more suitable for learning new associations and discovering new knowledge with highly likely relevance to a query. This work concludes that it is more suitable than Google for an exploratory search.

Document network is a kind of intriguing dataset which can provide both topical (textual content) and topological (relational link) information. A key point in viably modeling such datasets is to discover proper denominators beneath the two different types of data, text and link. Most previous work introduces the assumption that documents closely linked with each other share common latent topics. However, the heterophily (i.e., tendency to link to different others) of nodes is neglected, which is pervasive in social networks. In this paper, we simultaneously incorporate community detection and topic modeling in a unified framework, and appeal to Canonical Correlation Analysis (CCA) to capture the latent semantic correlations between the two heterogeneous latent factors, community and topic. Despite of the homophily (i.e., tendency to link to similar others) or heterophily, CCA can properly capture the inherent correlations which fit the dataset itself without any prior hypothesis. Logistic normal prior is also employed in modeling network to better capture the community correlations. We derive efficient inference and learning algorithms based on variational EM methods. The effectiveness of our proposed model is comprehensively verified on three different types of datasets which are namely hyperlinked networks of web pages, social networks of friends and coauthor networks of publications. Experimental results show that our approach achieves significant improvements on both topic modeling and community detection compared with the current state of the art. Meanwhile, our model is impressive in discovering correlations between extracted topics and communities.

The cost of powering servers, storage platforms and related cooling systems has become a major component of the operational costs in big data deployments. Hence, the design of energy-efficient Hadoop clusters has attracted significant research attentions in recent years. However, existing studies do not consider the impact of the complex interplay between workload and hardware heterogeneity on energy efficiency. In this paper, we find that heterogeneity-oblivious task assignment approaches are detrimental to both performance and energy efficiency of Hadoop clusters. Importantly, we make a counterintuitive observation that even heterogeneity-aware techniques that focus on reducing job completion time do not necessarily guarantee energy efficiency. We propose a heterogeneity-aware task assignment approach, E-Ant, that aims to minimize the overall energy consumption in a heterogeneous Hadoop cluster without sacrificing job performance. It adaptively schedules heterogeneous workloads on energy-efficient machines, without a priori knowledge of the workload properties. Furthermore, it provides the flexibility to trade off energy efficiency and job fairness in a Hadoop cluster. E-Ant employs an ant colony optimization approach that generates task assignment solutions based on the feedback of each task's energy consumption reported by Hadoop Task Trackers in an agile way. Experimental results on a heterogeneous cluster with varying hardware capabilities show that E-Ant improves the overall energy savings for a synthetic workload from Microsoft by 17% and 12% compared to Fair Scheduler and Tarazu, respectively.

Virtualizing Hadoop clusters provides many benefits, including rapid deployment, on-demand elasticity and secure multi-tenancy. However, a simple migration of Hadoop to a virtualized environment does not fully exploit these benefits. The dual role of a Hadoop worker, acting as both a compute node and a data node, makes it difficult to achieve efficient IO processing, maintain data locality, and exploit resource elasticity in the cloud. We find that decoupling per-node storage from its computation opens up opportunities for IO acceleration, locality improvement, and on-the-fly cluster resizing. To fully exploit these opportunities, we propose StoreApp, a shared storage appliance for virtual Hadoop worker nodes co-located on the same physical host. To completely separate storage from computation and prioritize IO processing, StoreApp pro-actively pushes intermediate data generated by map tasks to the storage node. StoreApp also implements late-binding task creation to take the advantage of prefetched data due to mis-aligned records. Experimental results show that StoreApp achieves up to 61% performance improvement compared to stock Hadoop and resizes the cluster to the (near) optimal degree of parallelism.

Urban traffic condition usually serves as a basic information for some intelligent urban applications, e.g., intelligent transportation system. But the acquisition of such information is often costly due to the dependency on equipments such as cameras and loop detectors. Crowdsensing can be utilized to gather vehicle-sensed data for traffic condition estimation. This way of data collection is economic. However, it has the problems of data uploading efficiency and data usage effectiveness. To deal with these problems, in this paper, we take into account the topology of the road net. We divide the road net into Road Sections and Junction Areas. Based on this division, we introduce a two-phased data collection and processing scheme named RTS (Road Topology based Scheme). It leverages the correlations among adjacent roads. In a junction area, data collected by vehicles is first processed and integrated by a sponsor vehicle. This sponsor vehicle will calculate the traffic condition locally. Both the selection of the sponsor and the calculation of the traffic condition utilize the road correlation. The sponsor then uploads the local data to a server. By employing the inherent relations among roads, the server processes data and estimates traffic condition for road sections unreached by vehicular data in a global vision. We conduct extensive experiments based on real vehicle trace data. The results indicate that, our design can commendably handle the problems of efficiency and effectiveness in the vehicular-crowdsensing-data based traffic condition evaluation.

As Hadoop is becoming increasingly popular in large-scale data analysis, there is a growing need for providing predictable services to users who have strict requirements on job completion times. While earliest deadline first scheduling (EDF) like algorithms are popular in guaranteeing job deadlines in real-time systems, they are not effective in a dynamic Hadoop environment, i.e., a Hadoop cluster with dynamically available resources. As there is a growing number of Hadoop clusters deployed on hybrid systems, e.g., infrastructure powered by mix of traditional and renewable energy, and cloud platforms hosting heterogeneous workloads, variable resource availability becomes common when running Hadoop jobs. In this paper, we propose, RDS, a Resource and Deadline-aware Hadoop job Scheduler that takes future resource availability into consideration when minimizing job deadline misses. We formulate the job scheduling problem as an online optimization problem and solve it using an efficient receding horizon control algorithm. To aid the control, we design a self-learning model to estimate job completion times and use a simple but effective model to predict future resource availability. We have implemented RDS in the open source Hadoop implementation and performed evaluations with various benchmark workloads. Experimental results show that RDS substantially reduces the penalty of deadline misses by at least 36% and 10% compared with Fair Scheduler and EDF scheduler, respectively.

In this paper, we mainly investigate capacity scaling laws of the mobile ad hoc social networks (MAHSNs)where social networking applications are implemented over the underlying mobile ad hoc networks. We model the real-world mobility pattern of mobile social users by introducing a clustered model that defines two levels of mobility, i.e., Strong mobility and weak mobility, according to the impacts of mobility on the gain of network capacity. To address the formation of social relationships among mobile social users, we adopt a distance and density aware social model called population-distance-based model that comprehensively and practically takes account of the clustering levels of friendship degree and distribution. Under those models, we derive the capacity scaling laws for social-broadcast sessions in MAHSNs. The results provide the exploratory insights into the impacts of users' mobility patterns and the formation of social relationships on the network capacity of MAHSNs.

Performance improvement and energy efficiency are two important goals in provisioning Internet services in datacenter servers. In this article, we propose and develop a self-tuning request batching mechanism to simultaneously achieve the two correlated goals. The batching mechanism increases the cache hit rate at the front-tier Web server, which provides the opportunity to improve an application’s performance and the energy efficiency of the server system. The core of the batching mechanism is a novel and practical two-layer control system that adaptively adjusts the batching interval and frequency states of CPUs according to the service level agreement and the workload characteristics. The batching control adopts a self-tuning fuzzy model predictive control approach for application performance improvement. The power control dynamically adjusts the frequency of Central Processing Units (CPUs) with Dynamic Voltage and Frequency Scaling (DVFS) in response to workload fluctuations for energy efficiency. A coordinator between the two control loops achieves the desired performance and energy efficiency. We further extend the self-tuning batching with DVFS approach from a single-server system to a multiserver system. It relies on a MIMO expert fuzzy control to adjust the CPU frequencies of multiple servers and coordinate the frequency states of CPUs at different tiers. We implement the mechanism in a test bed. Experimental results demonstrate that the new approach significantly improves the application performance in terms of the system throughput and average response time. At the same time, the results also illustrate the mechanism can reduce the energy consumption of a single-server system by 13% and a multiserver system by 11%, respectively.

With the globe facing various scales of natural disasters then and there, disaster recovery is one among the hottest research areas and the rescue and recovery services can be highly benefitted with the advancements of information and communications technology (ICT). Enhanced rescue effect can be achieved through the dynamic networking of people, systems and procedures. A seamless integration of these elements along with the service-oriented systems can satisfy the mission objectives with the maximum effect. In disaster management systems, services from multiple sources are usually integrated and composed into a usable format in order to effectively drive the decision-making process. Therefore, a novel service indexing method is required to effectively discover desirable services from the large-scale disaster service repositories, comprising a huge number of services. With this in mind, this paper presents a novel multilevel indexing algorithm based on the equivalence theory in order to achieve effective service discovery in large-scale disaster service repositories. The performance and efficiency of the proposed model have been evaluated by both theoretical analysis and practical experiments. The experimental results proved that the proposed algorithm is more efficient for service discovery and composition than existing inverted index methods.

Particle swarm optimization (PSO) algorithm is a population-based stochastic optimization technique. It is characterized by the collaborative search in which each particle is attracted toward the global best position (gbest) in the swarm and its own best position (pbest). However, all of particles' historical promising pbests in PSO are lost except their current pbests. In order to solve this problem, this paper proposes a novel composite PSO algorithm, called historical memory-based PSO (HMPSO), which uses an estimation of distribution algorithm to estimate and preserve the distribution information of particles' historical promising pbests. Each particle has three candidate positions, which are generated from the historical memory, particles' current pbests, and the swarm's gbest. Then the best candidate position is adopted. Experiments on 28 CEC2013 benchmark functions demonstrate the superiority of HMPSO over other algorithms.

This paper aims to design an efficient data forwarding scheme based on local activity and social similarity(LASS) for mobile social networks (MSNs). Various definitions of social similarity have been proposed as the criterion for relay selection, which results in various forwarding schemes. The appropriateness and practicality of various definitions determine the performances of these forwarding schemes. A popular definition has recently been proven to be more efficient than other existing ones, i.e., the more common interests between two nodes, the larger social similarity between them. In this work, we show that schemes based on such definition ignore the fact that members within the same community, i.e., with the same interest, usually have different levels of local activity, which will result in a low efficiency of data delivery. To address this, in this paper, we design a new data forwarding scheme for MSNs based on community detection in dynamic weighted networks, called Local-Activity and Social-Similarity, taking into account the difference of members' internal activity within each community, i.e., local activity. To the best of our knowledge, the proposed scheme is the first one that utilizes different levels of local activity within communities. Through extensive simulations, we demonstrate that LASS achieves better performance than state-of-the-art protocols.

In this paper, we investigate capacity scaling laws of wireless social networks under the social-based session formation. We model a wireless social network as a three-layered structure, consisting of the physical layer, social layer, and session layer; and we introduce a cross-layer distance & density-aware model, called the population-based formation model, under which: 1) for each node vk, the number of its friends/followers, denoted by qk, follows a Zipf's distribution with degree clustering exponent g; 2) qkanchor points are independently chosen according to a probability distribution with density function proportional to (Ek,X)-β, where Ek;Xis the expected number of nodes (population) within the distance |vk-X| to vk, and β is the clustering exponent of friendship formation; 3) finally, qknodes respectively nearest to those qkanchor points are selected as the friends of vk. We present the general density function of social relationship distribution, with general distribution of physical layer, serving as the basis for studying general capacity of wireless social networks. As the first step of addressing this issue, for the homogeneous physical layer, we derive the social-broadcast capacity under both generalized physical and protocol interference models, taking into account general clustering exponents of both friendship degree and friendship formation in a 2-dimensional parameter space, i.e., (γ,β) ϵ[0,∞)2. Importantly, we notice that the adopted model with homogenous physical layer does not sufficiently reflect the advantages of the population-based formation model in terms of realistic validity and practicability. Accordingly, we introduce a random network model, called the center-clustering random model (CCRM) with node distribution exponent δ ϵ [0, ∞), highlighting the clustering and inhomogeneity property in real-life networks, and discuss how to further derive more general network capacity over 3-dimensional parameter space (δ,γ,β) ϵ [0, ∞)3based on our results over (γ,β) ϵ [0, ∞)2.

As there are various risks of failure in its execution, a composite web service (CWS) requires a transactional mechanism to guarantee its reliable execution. Though the existing service selection methods have considered that its transactional properties may affect its quality of service (QoS) such as its execution time, some of these methods can just give the locally optimal transactional CWS while others can give globally optimal CWS only under a given fixed transactional workflow. This paper addresses the issue of selecting and composing web services via a genetic algorithm (GA) and gives a transaction and QoS-aware selection approach. First, it introduces transactional properties of a single web service and CWS and the transactional rules used to compose them. Next, it conducts the performance analysis of basic workflow patterns such as sequential, parallel, selectable, and loop patterns and develops an algorithm to compute the execution time of a complex CWS. Then, it presents a GA-based approach, which takes into account the execution time, price, transactional property, stability, and penalty-factor, to achieve globally optimal service selection. Finally, this paper reports experimental results that compare the proposed approach with the exhaustive search algorithm, transactional-QoS-driven selection algorithm, and transactional service selection algorithm. The experimental results show that the proposed algorithm is efficient and effective and can give a globally optimal transactional CWS.

van der Aalst et al. proved that the soundness problem is solvable in polynomial time for free-choice workflow nets (FCWF-nets). However, FCWF-nets cannot model most web services composition and interorganizational business processes because the interaction among processes does not usually satisfy the free-choice requirement. Asymmetricchoice workflow nets (ACWF-nets) as a larger class than FCWF-nets can model lots of such cases. Our previous work showed that the (weak) soundness problem is co-NP-hard for three-bounded ACWF-nets. Later, Tiplea et al. proved that for three-bounded acyclic ACWF-nets, the weak soundness problem is co-NP-complete. We sharp these results in this paper. First, we prove that for ACWF-nets, whether they are one-bounded or k-bounded (k > 1), the soundness problem is co-NP-hard. Second, it is proven that the soundness is equivalent to the weak soundness for any acyclic ACWF-nets, i.e., an acyclic ACWF-net is sound if and only if it is weakly sound.

In this paper, we study two tightly coupled issues, space-crossing community detection and its influence on data forwarding in mobile social networks (MSNs). We propose a communication framework containing the hybrid underlying network with access point (AP) support for data forwarding and the base stations for managing most of control traffic. The concept of physical proximity community can be extended to be one across the geographical space, because APs can facilitate the communication among long-distance nodes. Space-crossing communities are obtained by merging some pairs of physical proximity communities. Based on the space-crossing community, we define two cases of node local activity and use them as the input of inner product similarity measurement. We design a novel data forwarding algorithm Social Attraction and Infrastructure Support (SAIS), which applies similarity attraction to route to neighbor more similar to destination, and infrastructure support phase to route the message to other APs within common connected components. We evaluate our SAIS algorithm on real-life datasets from MIT Reality Mining and University of Illinois Movement (UIM). Results show that space-crossing community plays a positive role in data forwarding in MSNs. Based on this new type of community, SAIS achieves a better performance than existing popular social community-based data forwarding algorithms in practice, including Simbet, Bubble Rap and Nguyen's Routing algorithms.

In this paper, we study two tightly coupled issues: space-crossing community detection and its influence on data forwarding in Mobile Social Networks (MSNs) by taking the hybrid underlying networks with infrastructure support into consideration. The hybrid underlying network is composed of large numbers of mobile users and a small portion of Access Points (APs). Because APs can facilitate the communication among long-distance nodes, the concept of physical proximity community can be extended to be one across the geographical space. In this work, we first investigate a space-crossing community detection method for MSNs. Based on the detection results, we design a novel data forwarding algorithm SAAS (Social Attraction and AP Spreading), and show how to exploit the space-crossing communities to improve the data forwarding efficiency. We evaluate our SAAS algorithm on real-life data from MIT Reality Mining and University of Illinois Movement (UIM). Results show that space-crossing community plays a positive role in data forwarding in MSNs in terms of delivery ratio and delay. Based on this new type of community, SAAS achieves a better performance than existing social community-based data forwarding algorithms in practice, including Bubble Rap and Nguyen's Routing algorithms.

While major cloud service operators have taken various initiatives to operate their sustainable data enters with green energy, it is challenging to effectively utilize the green energy since its generation depends on dynamic natural conditions. Fortunately, the geographical distribution of data enters provides an opportunity for optimizing the system performance by distributing cloud workloads. In this paper, we propose a holistic heterogeneity-aware cloud workload placement and migration approach, sCloud, that aims to maximize the system good put in distributed self-sustainable data enters. sCloud adaptively places the transactional workload to distributed data enters, allocates the available resource to heterogeneous workloads in each data enter, and migrates batch jobs across data enters, while taking into account the green power availability and QoS requirements. We formulate the transactional workload placement as a constrained optimization problem that can be solved by nonlinear programming. Then, we propose a batch job migration algorithm to further improve the system good put when the green power supply varies widely at different locations. We have implemented sCloud in a university cloud test bed with real-world weather conditions and workload traces. Experimental results demonstrate sCloud can achieve near-to-optimal system performance while being resilient to dynamic power availability. It outperforms a heterogeneity-oblivious approach by 26% in improving system good put and 29% in reducing QoS violations.

In this paper, we model the data dissemination in online social networks (OSNs) and study the scaling laws of traffic load. We propose a three-layered system model to formulate data dissemination sessions for social applications in OSNs. The layered model consists of the physical network layer, social relationship layer, and application session layer. By analyzing mutual relevances among these three layers, we investigate the geographical distribution feature of dissemination sessions in OSNs. Based on this, we derive the traffic load of OSNs under a realistic assumption that every source sustains a data generating rate of constant order. To the best of our knowledge, this is the first work to address the issue of traffic load scaling for OSNs by modeling the social data dissemination from a layered perspective.

Load imbalance is a major source of overhead in Hadoop where the uneven distribution of input data among tasks can significantly delays the job completion. Running Hadoop in a private cloud opens up opportunities for mitigating data skew with elastic resource allocation, where stragglers are expedited with more resources, yet introduces problems that often cancel out the performance gain: (1) performance interference from co running jobs may create new stragglers, (2) there exist a semantic gap between Hadoop task management and resource pool-based virtual cluster management preventing efficient resource usage. We present Flex Slot, a user-transparent task slot management scheme that automatically identifies map stragglers and resizes their slots accordingly to accelerate task execution. Flex Slot adaptively changes the number of slots on each virtual node to promote efficient usage of resource pool. Experimental results with representative benchmarks show that Flex Slot effectively reduces job completion time by 46% and achieves better resource utilization.

A critical function of wireless sensor networks (WSNs) is data gathering. One is often only interested in collecting a specific function of the sensor measurements at a sink node, rather than downloading all the raw data from all the sensors. In this paper, we study the capacity of computing and transporting the specific functions of sensor measurements to the sink node, called aggregation capacity, for WSNs. We focus on random WSNs that can be classified into two types: random extended WSN and random dense WSN. All existing results about aggregation capacity are studied for dense WSNs, including random cases and arbitrary cases, under the protocol model (ProM) or physical model (PhyM). In this paper, we propose the first aggregation capacity scaling laws for random extended WSNs. We point out that unlike random dense WSNs, for random extended WSNs, the assumption made in ProM and PhyM that each successful transmission can sustain a constant rate is over-optimistic and unpractical due to transmit power limitation. We derive the first result on aggregation capacity for random extended WSNs under the generalized physical model. Particularly, we prove that, for the type-sensitive divisible perfectly compressible functions and type-threshold divisible perfectly compressible functions, the aggregation capacities for random extended WSNs with${\mbi {n}}$nodes are of order$\Thetab ({{{({\bf log} {\mbi {n}})}^{ - {\alphab \over {\bf 2}} - {\bf 1}}}})$and$\Theta ({{{{{({\bf log} {\mbi {n}})}^{ - \alphab /{\bf 2}}}} \over {{\bf log}{\bf log} {\mbi {n}}}}})$, respectively, where$\alphab \gt {\bf 2}$denotes the power attenuation exponent in the generalized physical model. Furthermore, we improve the aggregation throughput for general divisible perfectly compressible functions to$\Omegab ({{{({\bf log} {\mbi {n}})}^{ - {\alphab \over {\bf 2}}}}})$by choosing$\Thetab ({\bf log}\; {\mbi {n}})$sensors from a small region (relative to the whole region) as sink nodes.

In this paper, we focus on the asymptotic capacity and delay, and their tradeoffs in mobile ad hoc networks (MANETs). As we all know, some fixed rate communication models such as the protocol model and the physical model have been studied in the past. However, our work aims to investigate the impact of an adaptive rate communication model on capacity-delay tradeoffs in MANETs under classical mobility models. Specifically, we adopt a well-known adaptive rate model called the generalized physical model (GphyM). The mobility of nodes is characterized by two broad classes of practical mobility models and they are hybrid random walk models and discrete random direction models. The two models generalize many mobility models studied in the literature, including the random walk, i.i.d., Brownian, and random way point models. For each mobility model, we derive the optimal delay for the optimal per-session unicast capacity (that of constant order Θ(1)) under the generalized physical model, depending on the individual parameters of mobility models. In particular, we show that for the i.i.d. model, compared with those under the protocol and physical models, the adaptive feature of link rate under the generalized physical model results in a significant decrease in the optimal delay for the optimal capacity; more precisely, both the optimal capacity and optimal delay can be simultaneously achieved, while there is no improvement for the random way-point model.

Automated server parameter tuning is crucial to performance and availability of Internet applications hosted in cloud environments. It is challenging due to high dynamics and burstiness of workloads, multi-tier service architecture, and virtualized server infrastructure. In this paper, we investigate automated and agile server parameter tuning for maximizing effective throughput of multi-tier Internet applications. A recent study proposed a reinforcement learning based server parameter tuning approach for minimizing average response time of multi-tier applications. Reinforcement learning is a decision making process determining the parameter tuning direction based on trial-and-error, instead of quantitative values for agile parameter tuning. It relies on a predefined adjustment value for each tuning action. However it is nontrivial or even infeasible to find an optimal value under highly dynamic and bursty workloads. We design a neural fuzzy control based approach that combines the strengths of fast online learning and self-adaptiveness of neural networks and fuzzy control. Due to the model independence, it is robust to highly dynamic and bursty workloads. It is agile in server parameter tuning due to its quantitative control outputs. We implemented the new approach on a testbed of virtualized data center hosting RUBiS and WikiBench benchmark applications. Experimental results demonstrate that the new approach significantly outperforms the reinforcement learning based approach for both improving effective system throughput and minimizing average response time.

Internet data are heterogeneous, redundant, disordered, and exponentially growing. Finding the right information from them becomes an ever-challenging issue. Existing technologies such as inverted index and keyword matching can list user webpage matching with given search keywords. They cannot recognize potential relations among webpages to meet some rising user needs, e.g., exploratory search and personalized search. We propose an indexing network model that organizes information in webpages at three levels: words, webpages, and categories, thereby leading to a semantic association graph. Words are used as the description of webpages and categories. Webpage classification is used to gather similar webpages together. Hyperlinks imply the wisdom of the webpage creator, which can help us generate semantic relations among categories. With a clear organizational structure, an indexing network can provide support for many important applications including intelligent information retrieval, recommendation and decision support. In order to provide access to interfaces for the proposed indexing network, an indexing network algebra is defined. Finally, to validate the proposed model, an indexing network is generated based on 30 million webpages and its structure is analyzed. We also give methods to achieve “browsing navigation” and “personalized search” based on the generated network. Results reveal that the use of an indexing network can greatly facilitate exploratory information retrieval and personalized search.

The creation of value-added services by automatic composition of existing ones is gaining significant momentum as the potential silver bullet in service-oriented computing. A large number of composition methods have been proposed, and most of them are based on the matching of input and output parameters of services only. However, most services in the real world are not universally applicable, and some applicable conditions or restrictions are imposed on them by their providers. Such constraints have a great impact on service composition, but have been largely ignored by the existing methods. In this paper, they are discussed and defined, and a simple formal expression is adopted to describe them. Two novel concepts, called service intension and service extension, are presented, which allow one to divide the basic elements of a web service definition into two parts. Consequently, their use allows us to propose a constraint-aware service composition method in which service constraints are well taken care. The proposed solution includes a graph search-based algorithm and two novel preprocessing methods. A publicly available test set from ICEBE05 is used to evaluate and analyze the proposed methodology.

E-commerce and online shopping with a third-party payment platform have rapidly developed recently, and encountered many fault tolerance and security problems concerned by users. The causes of these problems include malicious behavior and imperfect business processes. The latter lead to the emergence of security vulnerabilities and loss of user funds which become more and more serious these years. We focus on the business process of e-commerce, and propose a formal model for constructing an e-commerce business process called an E-commerce Business Process Net. It integrates both data and control flows based on Petri nets. Rationality and transaction consistency are defined and validated to guarantee the transaction properties of an e-commerce business process. This paper offers a complete methodology for modeling and validating an e-commerce system with a third-party payment platform from the view point of a business process. Its use enables a designer to identify errors early in the design process and correct them before the deployment phase. In order to demonstrate the applicability and feasibility of the methodology, we have modeled and validated a real-world e-commerce business process and discovered the problems that cause the violation of transaction properties.

Owing to their rapid development, services are increasing rapidly in quantity. The consequence is that there are so many services that share the same or similar functions. Therefore, it is important to select a credible and optimal service. Reputation as one of the important parameters of services plays a significant role in the decision support for service selection. This paper proposes a novel two-phase method to calculate service reputation. The first phase uses a dynamic weight formula to calculate reputation such that it can reflect the latest tendency of a service. The second one uses an olfactory response formula to mitigate the negative effect of unfair ratings. Some experiments are conducted and the results validate the effectiveness of the proposed method.

A type of refinement operations of time Petri nets is presented for design, analysis and verification of complex real-time systems. First, the behavior preservation is studied under time constraints in a refinement operation, and a sufficient condition for behavior preservation is obtained. Then, the property preservation is considered, and the results indicate that if the refinement operation of time Petri nets satisfies behavior preservation, it can also preserve properties such as boundedness and liveness. Finally, based on the behavior preservation, a reachability decidability algorithm of a refined time Petri net is designed using the reachability trees of its original net and subnet. The research results are illustrated by an example of designing, analyzing and verifying a real-time manufacturing system.

We study the capacity scaling laws for the cognitive network that consists of the primary hybrid network (PhN) and secondary ad hoc network (SaN). PhN is further comprised of an ad hoc network and a base station-based (BS-based) network. SaN and PhN are overlapping in the same deployment region, operate on the same spectrum, but are independent with each other in terms of communication requirements. The primary users (PUs), i.e., the ad hoc nodes in PhN, have the priority to access the spectrum. The secondary users (SUs), i.e., the ad hoc nodes in SaN, are equipped with cognitive radios, and have the functionalities to sense the idle spectrum and obtain the necessary information of primary nodes in PhN. We assume that PhN adopts one out of three classical types of strategies, i.e., pure ad hoc strategy, BS-based strategy, and hybrid strategy. We aim to directly derive multicast capacity for SaN to unify the unicast and broadcast capacities under two basic principles: 1) The throughput for PhN cannot be undermined in order sense due to the presence of SaN. 2) The protocol adopted by PhN does not alter in the interest of SaN, anyway. Depending on which type of strategy is adopted in PhN, we design the optimal-throughput strategy for SaN. We show that there exists a threshold of the density of SUs according to the density of PUs beyond which it can be proven that: 1) when PhN adopts the pure ad hoc strategy or hybrid strategy, SaN can achieve the multicast capacity of the same order as it is stand-alone; 2) when PhN adopts the BS-based strategy, SaN can asymptotically achieve the multicast capacity of the same order as if PhN were absent, if some specific conditions in terms of relations among the numbers of SUs, PUs, the destinations of each multicast session in SaN, and BSs in PhN hold.

Such concurrent systems as Web services and workflow systems can be viewed as a composition of a set of subsystems. Subsystems interact with each other through a set of message channels in order to perform a task. This work defines a class of Petri nets called interactive Petri nets (IPNs) to model these systems. IPNs can be used to analyze their behavior, find potential problems, and then improve their designs. Compatibility is an important concept for a composed system and reflects the possibility of correct/proper interaction among its subsystems. In order to characterize different cooperative abilities in practice, compatibility and weak compatibility are defined for IPNs. Some relationships among (weak) compatibility, liveness, reversibility, and boundedness are revealed. Based on them, this work proves that the (weak) compatibility problem is co-NP-hard. A taxonomy is also presented for IPNs in order to explore whether some subclasses can be analyzed efficiently. Based on it, we can identify several IPN subclasses, some of which can be analyzed in polynomial time.

The use of information technology and management systems for the betterment of health care is more and more important and popular. However, existing efforts mainly focus on informatization of hospitals or medical institutions within the organizations, and few are directly oriented to the patients, their families, and other ordinary people. The strong demand for various medical and public health care services from customers calls for the creation of powerful individual-oriented personalized health care service systems. Service computing and related technologies can greatly help one in fulfilling this task. In this paper, we present PHISP: a Public-oriented Health care Information Service Platform, which is based on such technologies. It can support numerous health care tasks, provide individuals with many intelligent and personalized services, and support basic remote health care and guardianship. In order to realize the personalized customization and active recommendation of intelligent services for individuals, several key techniques for service composition are integrated, which can support branch and parallel control structures in the process models of composite services and are highlighted in this paper.

With the rapid development of service-oriented computing (SOC) and service-oriented architecture (SOA), the number of services is rapidly increasing. How to organize and manage services effectively in repositories to improve the efficiency of service discovery and composition is important. This paper proposes three categorization rules to classify services for a large scale repository to form a relational taxonomy. The service retrieve scope can be drastically narrowed by this taxonomy. Therefore, the efficiency of service discovery and service composition can be greatly improved. We evaluate and compare the performance of the proposed method and other related ones via a publicly available test set, ICEBE05. The experimental results validate the effectiveness and high efficiency of the proposed one.

Extensive efforts have been made to study the asymptotic capacity, delay, and their tradeoffs for large-scale mobile ad hoc networks, under different mobility models and communication models. Majority results adopt the fixed-rate communication model, such as the protocol model and physical model, and none of them breaks the limitation of tradeoffs: delay/capacity = ω(1) so far, even for the simplest i.i.d. model. In this work, we investigate this problem under the Gaussian channel model, and demonstrate that the delay-capacity tradeoffs can be further improved by designing new two-hop strategy under a general mobility model, called hybrid random walk mobility model (HRWMM). We found that the capacity and delay have several regions, depending on the freedom degree γ ϵ [0, 1] of mobile nodes. Specifically, we show that under the prerequisite of ensuring the optimal per-session capacity, i.e., of order Θ(1): (1) for 0 <; γ ≤ 1, the optimal delay under the Gaussian channel model is smaller than that under the protocol model or physical model; (2) for γ = 0, i.e., ordinary random walk model, it is not larger than the delay under the protocol model or physical model; (3) for γ = 1, i.e., i.i.d. model, the capacity and delay can simultaneously achieve the optimal order, i.e., Θ(1).

We study the asymptotic networking-theoretic multicast capacity bounds for random extended networks (REN) under Gaussian channel model, in which all wireless nodes are individually power-constrained. During the transmission, the power decays along path with attenuation exponent \alpha > 2. In REN, n nodes are randomly distributed in the square region of side length \sqrt{n}. There are n_s randomly and independently chosen multicast sessions. Each multicast session has n_d+1 randomly chosen terminals, including one source and n_d destinations. By effectively combining two types of routing and scheduling strategies, we analyze the asymptotic achievable throughput for all n_s=\omega (1) and n_d. As a special case of our results, we show that for n_s=\Theta (n), the per-session multicast capacity for REN is of order \Theta ({1\over \sqrt{n_d n}}) when n_d=O({n\over ({\log n})^{\alpha +1}} ) and is of order \Theta ({1\over n_d} \cdot (\log n)^{-{\alpha \over 2} }) when n_d=\Omega ({n\over \log n} ).

In this paper, we study multicast capacity for cognitive networks. We consider the cognitive network model consisting of two overlapping ad hoc networks, called the primary ad hoc network (PaN) and secondary ad hoc network (SaN), respectively. PaN and SaN operate on the same space and spectrum. For PaN (or SaN, respectively), we assume that primary (or secondary, respectively) nodes are placed according to a Poisson point process of intensity n (or m, respectively) over a unit square region. We randomly choose n_s (or m_s, respectively) nodes as the sources of multicast sessions in PaN (or SaN, respectively), and for each primary source v^p (or secondary source v^s, respectively), we pick uniformly at random n_d primary nodes (or m_d secondary nodes, respectively) as the destinations of v^p (or v^s, respectively). Above all, we assume that PaN can adopt the optimal protocol in terms of the throughput. Our main work is to design the multicast strategy for SaN by which the optimal throughput can be achieved, without any negative impact on the throughput for PaN in order sense. Depending on n_d and n, we choose the optimal one for PaN from two strategies called percolation strategy and connectivity strategy, respectively. Subsequently, we design the corresponding throughput-optimal strategy for SaN. We derive the regimes in terms of n, n_d, m, and m_d in which the upper bounds on multicast capacities for PaN and SaN can be achieved simultaneously. Unicast and broadcast capacities for the cognitive network can be derived by our results as the special cases by letting n_d=1 (or m_d=1) and n_d=n-1 (or m_d=m-1), respectively, which enhances the generality of this work.

In this work, for a wireless sensor network (WSN) of n randomly placed sensors with node density \lambda \in [1,n], we study the tradeoffs between the aggregation throughput and gathering efficiency. The gathering efficiency refers to the ratio of the number of the sensors whose data have been gathered to the total number of sensors. Specifically, we design two efficient aggregation schemes, called single-hop-length (SHL) scheme and multiple-hop-length (MHL) scheme. By novelly integrating these two schemes, we theoretically prove that our protocol achieves the optimal tradeoffs, and derive the optimal aggregation throughput depending on a given threshold value (lower bound) on gathering efficiency. Particularly, we show that under the MHL scheme, for a practically important set of symmetric functions called divisible perfectly compressible (DPC) functions, including the mean, max, and various kinds of indicator functions, etc., the data from \Theta (n) sensors can be aggregated to the sink at the throughput of a constant order \Theta (1), implying that, our MHL scheme is indeed scalable.

This paper presents a class of Petri nets, process nets with channels (PNCs) that can model some types of concurrent systems in two aspects: process and interaction. Its significance lies in offering efficient analysis and verification methods for these systems. PNCs belong to the class of extended free choice nets. This paper establishes the conditions to examine their liveness, reversibility, and reachability based on their structural characteristics. Siphons, traps, and a state equation are used to describe these conditions such that analysis techniques based on reachability graphs and siphon enumeration are avoided. A polynomial-time algorithm is presented for the liveness analysis, and an effective method is also given to decide the reachability. A real-world example is used to illustrate the application of PNCs.

The information technology has been recognized as one of the most important means to improve health care and curb its ever-increasing cost. However, existing efforts mainly focus on informatization of hospitals or medical institutions within organizations, and few are directly oriented to individuals. The strong demand for various health services from customers calls for the creation of powerful individual-oriented personalized health care service systems. Web service composition (WSC) and related technologies can greatly help one build such systems. This paper aims to present a newly developed platform called a Public oriented Health care Information Service Platform (PHISP) and several novel WSC techniques that are used to build it. Among them include WSC techniques that can well support branch and parallel structures.

A critical function of wireless sensor networks (WSNs) is data gathering. While, one is often only interested in collecting a relevant function of the sensor measurements at a sink node, rather than downloading all the data from all the sensors. This paper studies the capacity of computing and transporting the specific functions of sensor measurements to the sink node, called aggregation capacity, for WSNs. It focuses on random WSNs that can be classified into two types: random extended WSN and random dense WSN. All existing results about aggregation capacity are studied for dense WSNs, including random cases and arbitrary cases, under the protocol model (ProM) or physical model (PhyM). In this paper, we propose the first aggregation capacity scaling laws for random extended WSNs. We point out that unlike random dense WSNs, for random extended WSNs, the assumption made in ProM and PhyM that each successful transmission can sustain a constant rate is over-optimistic and unpractical due to transmit power limitation.We derive the first result on aggregation capacity for random extended WSNs under the generalized physical model. Particularly, we prove that, for the type-sensitive perfectly compressible functions and type-threshold perfectly compressible functions, the aggregation capacities for random extended WSNs with n nodes are of order Θ ((log n)-β/2-1) and Θ (((log n)-β/2)/(log log n)), respectively, where β >; 2 denotes the power attenuation exponent in the generalized physical model.

We study the general scaling laws of the capacity for random wireless networks under the generalized physical model. The generality of this work is embodied in three dimensions denoted by (λ ∈ [1, n], nd∈ [1, n], ns∈ (1, n]). It means that: (1) We study the random network of a general node density λ ∈ [1, n], rather than only study either random dense network (RDN, λ = n) or random extended network (REN, λ = 1) as in the literature. (2) We focus on the multicast capacity to unify unicast and broadcast capacities by setting the number of destinations for each session as a general value nd∈ [1, n]. (3)We allow the number of sessions changing in the range ns∈ (1, n], rather than assume that ns= Θ(n) as in the literature.We derive the general lower bounds on the capacity for the arbitrary case of (λ, nd, ns). Particularly, we show that for the special cases (λ = 1, nd∈ [1, n], ns= n) and (λ = n, nd∈ [1, n], ns= n), our schemes achieve the highest multicast throughputs proposed in the existing works.

In this work, for a wireless sensor network (WSN) of n randomly placed sensors with node density λ ∈ [1, n], we study the tradeoffs between the aggregation throughput and gathering efficiency. The gathering efficiency refers to the ratio of the number of the sensors whose data have been gathered to the total number of sensors. Specifically, we design two efficient aggregation schemes, called single-hop-length (SLH) scheme and multiple-hop-length (MLH) scheme. By novelly integrating these two schemes, we theoretically prove that our protocol achieves the optimal tradeoffs, and derive the optimal aggregation throughput depending on a given threshold value (lower bound) on gathering efficiency. Particularly, we show that under the MLH scheme, for a practically important set of symmetric functions called perfectly compressible functions, including the mean, max, or various kinds of indicator functions, etc., the data from Θ(n) sensors can be aggregated to the sink at the throughput of a constant order Θ(1), implying that our MLH scheme is indeed scalable.

Radio frequency identification (RFID) is a technology where a reader device can "sense'' the presence of a close by object by reading a tag device attached to the object. To guarantee the coverage quality, multiple RFID readers can be deployed in the given region. In this paper, we consider the problem of activation schedule for readers in a multi-reader environment. In particular, we try to design a schedule for readers to maximize the number of served tags per time-slot while avoiding various interferences. We first develop a centralized algorithm under the assumption that different readers may have different interference and interrogation radius. Next, we propose a novel algorithm which does not need any location information of the readers. Finally, we extend the previous algorithm in distributed manner in order to suit the case where no central entity exists. We conduct extensive simulations to study the performances of our proposed algorithm. And our evaluation results corroborate our theoretical analysis.

This paper gives a necessary and sufficient condition for the liveness of normal nets, i.e. a normal net with a given initial marking is live if and only if it is structurally repetitive and each minimal siphon is marked in any reachable marking. Furthermore, it is proved that a normal net is structurally live if and only if it is structurally repetitive. Finally, we prove that a weakly persistent net, which is a special normal net, is live for a given initial marking if and only if it is structurally repetitive and each minimal siphon is marked in the initial marking. That is to say, the liveness of weakly persistent nets can be decided by the net structure and the initial marking only.

We study the multicast capacity for hybrid wireless networks consisting of ordinary ad hoc nodes and base stations under Gaussian Channel model, which generalizes both the unicast and broadcast capacities for hybrid wireless networks. Assume that all ordinary ad hoc nodes transmit at a constant power P, and the power decays along the path, with attenuation exponent \alpha >2. The data rate of a transmission is determined by the Signal to Interference plus Noise Ratio (SINR) at the receiver as B \log (1 + {\rm SINR}). The ordinary ad hoc nodes are placed in the square region {\cal A}(a) of area a according to a Poisson point process of intensity n/a. Then, m additional base stations (BSs) acting as the relaying communication gateways are placed regularly in the region {\cal A}(a ), and are connected by a high-bandwidth wired network. Let a=n and a=1, we construct the hybrid extended network (HEN) and hybrid dense network (HDN), respectively. We choose randomly and independently n_s ordinary ad hoc nodes to be the sources of multicast sessions. We assume that each multicast session has n_d randomly chosen terminals. Three broad categories of multicast strategies are proposed. The first one is the hybrid strategy, i.e., the multihop scheme with BS-supported, which further consists of two types of strategies called connectivity strategy and percolation strategy, respectively. The second one is the ordinary ad hoc strategy, i.e., the multihop scheme without any BS-supported. The third one is the classical BS-based strategy under which any communication between two ordinary ad hoc nodes is relayed by some specific BSs. According to the different scenarios in terms of m, n, and n_d, we select the optimal scheme from the three categories of strategies, and derive the achievable multicast throughput based on the optimal decision.

In this work, we present a monitor and rescue system utilizing hybrid networks which is a integration of stationary sensor networks and mobile sensor networks: stationary sensor networks comprised of large numbers of small, simple, and inexpensive wireless sensors, and the mobile sensor network contains a set of mobile sensors (robots). The static sensors in our network have “monitoring” ability, i.e., any activated static sensor can detect the event as long as its sensing range intersects the event region. And the mobile sensors have “moving” and “rescuing” ability, e.g., they can move toward the event region with limited speed and further perform certain rescuing/processing operations on the event. We can consider the event as a hazard, e.g., wild fire, and the mobile sensors as fireman robots. As soon as the fire is detected by the static sensors, the fireman robots are expected to move from its initial location to the hazard region within minimum latency. We define the reaction delay of the system as the delay from the occurrence of event till at least one mobile sensor reaches the event. In order to satisfy certain reaction delay requirement while minimizing the total cost, we propose a number of deployment strategies for the stationary sensor network and mobile sensor network respectively. We further design a random wake-up scheduling for the static sensors for the sake of energy efficiency. Finally, we propose a pure distributed motion strategy for mobile sensors without reliance on localization services such as GPS, focusing on simple algorithms for distributed decision making and information propagation. We demonstrate the efficacy of our system in simulation, providing empirical results.

We study the capacity scaling laws for the cognitive network that consists of the primary hybrid network (PhN) and secondary ad hoc network (SaN). PhN is further comprised of an ad hoc network and a base station based (BS-based) network. SaN and PhN are overlapping in the same deployment region, operate on the same spectrum, but are independent with each other in terms of communication requirements. The primary users (PUs), i.e., the ad hoc nodes in PhN, have the priority to access the spectrum. The secondary users (SUs), i.e., the ad hoc nodes in SaN, are equipped with cognitive radios, and have the functionalities to sense the idle spectrum and obtain the necessary information of primary nodes in PhN. We assume that PhN adopts one out of three classical types of strategies, i.e., pure ad hoc strategy, BS-based strategy, and hybrid strategy. We aim to directly derive multicast capacity for SaN to unify the unicast and broadcast capacity under two basic principles: (1) The throughput for PhN cannot be undermined in order sense due to the presence of SaN. (2) The protocol adopted by PhN does not alter in the interest of SaN, anyway. Depending on which type of strategy is adopted in PhN, we design the optimal-throughput strategy for SaN. We show that there exists a threshold of the density of SUs according to the density of PUs beyond which it can be proven that: (1) when PhN adopts the pure ad hoc strategy or hybrid strategy, SaN can achieve the multicast capacity of the same order as it is stand-alone; (2) when PhN adopts the BS-based strategy, SaN can asymptotically achieve the multicast capacity of the same order as if PhN were absent, if some conditions of the relations among the number of SUs, PUs, the destinations of each multicast sessions in SaN, and the base stations in PhN hold.

This paper proposes the concept of Key-resource/operation-place Pairs (KP) of S3PR (systems of simple sequential processes with resources). Based on KP, two policies are presented to prevent deadlocks in flexible manufacturing systems (FMS) that can be modeled by S3PR. The idea is to control some key resource places only to guarantee that all strict minimal siphons (SMS) never become empty, thereby making the controlled system live. It enables one to design two easy-to-implement control policies. The first one can guarantee that the controlled system is live, and the second one can also make the controlled system live if there is no SMS containing any control place in the controlled system. At last, a well-known FMS example is used to illustrate the proposed concept and policies.

We study the multicast capacity for hybrid wireless networks consisting of ordinary wireless nodes and base stations under Gaussian Channel model, which generalizes both the unicast capacity and broadcast capacity for hybrid wireless networks. We simply consider the hybrid extended network, where the ordinary wireless nodes are placed in the square region An with side-length sqrt n according to a Poisson point process with unit intensity. In addition, $m$ additional base stations (BSs) serving as the relay gateway are placed regularly in the region An and they are connected by a high-bandwidth wired network. Three broad categories of multicast strategies are proposed in this paper. According to the different scenarios in terms of m, n and n_d, we select the optimal scheme from the three categories of strategies, and derive the achievable multicast throughput based on the optimal decision.

We focus on the networking-theoretic multicast capacity for both random extended networks (REN) and random dense networks (RDN) under Gaussian Channel model, when all nodes are individually power-constrained. During the transmission, the power decays along path with the attenuation exponent alpha > 2. In REN and RDN, n nodes are randomly distributed in the square region with side-length radic(n) and 1, respectively. We randomly choose nsnodes as the sources of multicast sessions, and for each source v, we pick uniformly at random ndnodes as the destination nodes. Based on percolation theory, we propose multicast schemes and analyze the achievable throughput by considering all possible values of nsand nd. As a special case of our results, we show that for ns= Theta(n), the per-session multicast capacity of RDN is Theta((1)/(radic(ndn))) when nd= O((n)/((log n)3)) and is Theta((1)/(n)) when nd= Omega((1)/(log n)); the per-session multicast capacity of REN is Theta((1)/radic(ndn)) when nd= O((n)/((log n)alpha+1)) and is Theta((1)/(nd) ldr (log n)-(alpha)/(2)) when nd= Omega((n)/(log n)).

In this paper, we study the capacity of cognitive networks. We focus on the network model consisting of two overlapping ad hoc networks, called the primary ad hoc network (PaN) and secondary ad hoc network (SaN), respectively. PaN and SaN operate on the same space and spectrum. For PaN (or SaN resp.) we assume that primary (or secondary resp.) nodes are placed according to a Poisson point process of intensity n (or m resp.) over a unit square region. We randomly choose ns(or msresp.) nodes as the sources of multicast sessions in PaN (or SaN resp.), and for each primary source vp(or secondary source vs), we pick uniformly at random ndprimary nodes (or mdsecondary nodes) as the destinations of vp(or vs). Above all, we assume that PaN can adopt the optimal protocol in terms of the throughput. Our main work is to design the multicast strategy for SaN by which it can achieve the optimal throughput, without any negative impact on the throughput for PaN in order sense. Specifically, depending on ndand n, we choose the optimal strategy for PaN from two candidates called percolation strategy and connectivity strategy, respectively. Subsequently, we design the corresponding throughput-optimal strategy for SaN. We further derive the regimes for n, nd, m and mdwhere the throughputs for PaN and SaN can simultaneously achieve the upper bound of their capacities asymptotically.

We mainly study the achievable multicast throughput (AMT) for homogeneous wireless ad hoc networks under Gaussian channel model. We focus on two typical random networks, i.e., random extended networks (REN) and random dense networks (RDN). In REN and RDN, n nodes are randomly distributed in the square region with side-length n1/2and 1, respectively. We randomly choose nsnodes as the sources of multicast sessions, and for each source v, we pick uniformly at random ndnodes as the destinations. We propose multicast schemes without using percolation theory, and analyze the achievable multicast throughput by taking account of all possible values of nsand nd. As a special case of our results, we show that for ns=Θ(n), under specified conditions.

In cooperative systems (CSs), participants cannot usually ensure the correct behavior of their partners. Obligations and proofs of participants have to be performed together to achieve a common goal in a real cooperation. Without adequate accountability assurances of actions, there is no means of reliably enforcing punitive measures against fraudulent participants. However, the existing formal methods for analyzing CSs cannot properly deal with accountability and obligations. As such, this paper proposes a new class of labeled Petri net (LPN) models. The behavior of each partner is represented by an LPN, while a CS is modeled by the combination of all partners' LPN models. The behavioral properties of an overall modeled system can be well verified only by analyzing each individual LPN. LPNs provide the integration of formal notations with graphical notations and formal proofs with commonly used verification techniques. The obligations are verified based on LPN languages and the nonblocking properties of action sequences, while accountability can be proved by the network conditions and local action sequences on each partner's side. The proposed approaches are illustrated with the modeling and analysis of a purchase transaction using the Internet Open Trading Protocol.

A deadlock-checking approach for one-place unbounded Petri nets is presented based on modified reachability trees (MRTs). An MRT can provide some useful information that is lost in a finite reachability tree, owing to MRT's use of the expression a + bn(i) rather than symbol omega to represent the value of the components of a marking. The information is helpful to property analysis of unbounded Petri nets. For the deadlock-checking purpose, this correspondence paper classifies full conditional nodes in MRT into two types: true and fake ones. Then, an algorithm is proposed to determine whether a full conditional node is true or not. Finally, a necessary and sufficient condition of deadlocks is presented. Examples are given to illustrate the method.

The current stepwise refinement operation of Petri nets mainly concentrates on property preservation, which is an effective way to analyze and verify complex systems. Further steps into this field are needed from the perspective of system synthesis and language preservation. First, the refinement of Petri nets is introduced based on a$k$-well-behaved Petri net, in which$k$tokens can be processed. Then, according to the different compositions of subsystems, well-, under- and overmatched refined Petri nets are proposed. In addition, the language and property relationships among sub-, original, and refined nets are studied to demonstrate behavior characteristics and property preservation in a system synthesis process. A manufacturing system is given as an example to illustrate the effectiveness of the proposed approach in synthesizing and analyzing the Petri nets of complex systems.

This paper shows how temporal Petri nets (TPNs) can be used to specify and analyze an Internet stock trading system. The dynamical behavior of the system and causality between events can be explicitly described by temporal formulas. The functional correctness of the modeled system is formally verified by using the inferential rules in temporal logic. Important properties of the system are analyzed based on its TPN model such as liveness, eventuality, and fairness properties. This paper demonstrates that TPNs can provide significant advantages in the design and analysis of business processes.

This paper introduces an automatic Web service composition method based on logical inference of Horn clauses in Petri net models. The available services and user request described in SA WSDL are translated into a set of Horn clauses and the composability rules of the services' input/output parameters are established using ontology reasoning. We choose Petri net as the model of this set of Horn clauses. T-invariant method of Petri nets is used to determine the existence of composite Web services that can fulfill user's requirement.

The existing formal techniques are not suitable for elegantly modeling passing value indeterminacy and describing batch processing function in real-time cooperative systems. Moreover, the correct behavior of the systems depends on not only the logical correctness of the results obtained through running workflows but also the time of producing them before critical deadlines. For these purposes, this paper proposes an interorganizational logical workflow net (ILWN) for modeling and analyzing real-time cooperative systems based on time Petri nets, workflow techniques, and temporal logic. Through attaching logical expressions to some actions of an ILWN model, the size of the model is reduced. Thus, ILWNs can efficiently mitigate the state explosion problem to some extent. Also, this paper analyzes the soundness of a subclass of ILWNs: the or-restricted ILWNs. A rigorous analysis approach is given based on their static net structures only. The concepts and techniques proposed in this paper are illustrated with a seller-buyer example in electronic commerce.

In this paper a genetic algorithm-based optimization procedure for ontology matching problem is presented as a feature-matching process. First, from a global view, we model the problem of ontology matching as an optimization problem of a mapping between two compared ontologies, and every ontology has its associated feature sets. Second, as a powerful heuristic search strategy, genetic algorithm is employed for the ontology matching problem. Given a certain mapping as optimizing object for GA, fitness function is defined as a global similarity measure function between two ontologies based on feature sets. Finally, a set of experiments are conducted to analysis and evaluate the performance of GA in solving ontology matching problem

Refinement operation of workflow nets is provided in this paper for modeling and analyzing integrated workflow. Structure, dynamic properties and behavior expression of refined workflow net are also discussed. These works proves that refinement operation with step-by-step refinement of transitions could realize hierarchical modeling of workflow as well as composite modeling of workflow integration Furthermore, the refinement operation can reduce complexity of model analysis. In fact the reliable refined nets satisfy soundness, and dynamic behavior of refined Petri nets was consistent with of original nets and subnets. Therefore the properties analysis and verification of refined workflow nets can be realized by properties of subnets using refinement operation. Moreover, the research results are successfully applied to designing, modeling and verification of layered workflows and their integration in Shanghai Traffic Information Service Application Grid (STISAG).