ABSTRACT Credit card fraud detection remains a challenging research problem due to the class imbalance issue caused by the rarity of fraudulent transactions. Classical oversampling techniques such as SMOTE, ADASYN and their variants help balance data but do not reflect the nonlinear structure of real‐world fraud, leading to poor generalization. Recent state‐of‐the‐art hybrid frameworks that combine deep generative models and ensemble learning improve performance but treat representation learning, augmentation and fusion as disconnected stages. To address these limitations, we propose a unified multistage framework that integrates representation learning, generative augmentation and intelligent ensemble fusion. Our framework first extracts autoencoder‐based latent representations to capture discriminative and interpretable features; then, a label‐conditioned VAE‐GAN uses these embeddings to generate realistic synthetic fraud samples; finally, the enriched features are projected into a fusion space and classified using a pool of diverse learners, whose outputs are consolidated through an embedding‐aware intelligent ensemble and a meta‐ensemble layer. We benchmark the framework against two categories of baselines: oversampling‐based methods and state‐of‐the‐art hybrid fraud detection systems. Experiments on the European cardholder dataset show that our approach achieves a macro F1‐score of 95.15% and balanced accuracy of 92.85%, outperforming both baseline categories by 2.8%. Additional experiments on the IEEE‐CIS Fraud Detection dataset further validate the generalizability of the proposed framework on large‐scale, heterogeneous and feature‐rich fraud data. The results demonstrate that the proposed framework not only improves detection accuracy under severe imbalances but also maintains interpretability, offering a robust and scalable foundation for reliable financial risk control.

Abstract Language models often underperform when fine tuned with limited labelled data. To reduce dependence on costly annotation, recent studies have explored clustering and pseudo labelling to leverage unlabeled text with only a few labelled examples for fine tuning. However, these methods face persistent challenges: clustering errors and pseudo-label mismatches can degrade fine-tuning performance. We propose Cluster-SCP , a novel framework that improves intra-cluster coherence and reduces cluster quantity to generate more reliable pseudo labels. The framework begins with K-Means initialization and applies an iterative refinement process, Intermediate Pseudo Labels , comprising two stages: embedding reassignment to minimize clustering errors, and cluster merging via contrastive learning with graph readouts. The refined pseudo labels are first used to fine tune language model, which is subsequently refined using a small set of labels. Experiments on three benchmark datasets show that Cluster-SCP consistently outperforms baseline and state-of-the-art methods, achieving at least a 5.3% accuracy improvement while reducing the number of clusters by 22%.

Recently, Large Language Models with Retrieval-Augmented Generation (RAG) have recently emerged as a powerful paradigm for sequential recommendation. However, existing methods typically retrieve items for each user without any principled mechanism for guaranteeing the reliability of generated recommendations, limiting their trustworthiness. To address this, we introduce SarRec : Statistically-guaranteed Augmented Retrieval for Recommendations, a framework that uses a simple retrieval step to provide relevant context and delivers calibrated, uncertainty-aware predictions with formal statistical guarantees. Specifically, SarRec first constructs the user's context set, utilizing a lightweight differentiable retrieval mechanism for identifying relevant context, and then calibrates the LLM's outputs by adapting the conformal prediction mechanism. We further provide a theoretical analysis that establishes an upper bound on the expected risk of recommendation performance metrics. Extensive experiments on multiple datasets from different domains validate the effectiveness of our framework.

Achieving reliable indoor 5G coverage in high-rise urban environments is challenging due to the significant signal attenuation of higher frequency bands through building materials and the intricate layouts of modern architecture. Traditionally, assessing in-building coverage to meet mobile network operators’ standards requires conducting indoor walk tests, a method that is both costly and time-consuming. This study aims to develop a scalable and cost-effective method for predicting and visualizing in-building 5G coverage using externally collected radio frequency (RF) data. We propose a methodology that employs AIoT-enabled drones equipped with cellular network scanners to gather key signal metrics, such as Received Signal Strength Indicator (RSSI) and Channel Quality Indicator (CQI), from the dominant 5G macrocell outside the building. The proposed Location-Aware Deep Neural Network (LA-DNN) integrates external radio frequency (RF) measurements with spatial coordinates, free-space path loss models, and material-specific attenuation factors to predict indoor RSSI and CQI. The model demonstrates robust predictive performance, achieving an R 2 of 0.9567 with a root mean squared error (RMSE) of 8.75 dBm for RSSI, and an R 2 of 0.9713 with an RMSE of 2.08 for CQI, validated against indoor walk-test data. Additionally, three-dimensional (3D) heatmaps are generated to visualize predicted coverage across building floors, facilitating the identification of coverage gaps and RF null zones. By minimizing the need for indoor measurements during initial in-building coverage assessments, the proposed framework significantly enhances the efficiency of site survey teams before implementing In-Building Coverage (IBC) systems.

Applying large vision models (LVMs) and large language models (LLMs) for item embedding is becoming cutting-edge for sequential recommendation, given their success in broad applications. Despite their advantages over traditional approaches, these models suffer more significant performance degradation on tail items against conventional ID-based solutions, which are largely overlooked by recent research. In this paper, we substantiate the above challenges as (1) all-in ground-truth, i.e., the standard cross-entropy (CE) loss focuses solely on the target items while treating all non-ground-truth equally, causing insufficient optimization for tail items, and (2) knowledge transfer tax, i.e., the knowledge encapsulated in LLMs and LVMs dominates the optimization process due to insufficient training for tail items. We propose Rewarding and reembedding, a simple yet efficient method to address the above challenges. Specifically, we reinitialize tail item embedding via a Gaussian distribution to alleviate knowledge transfer tax; besides, a rewarding function is incorporated in the CE loss, which adaptively adjusts item rewards during training to encourage the model to pay more attention to tail items rather than exclusively optimizing for ground-truth. Overall, our method enables a more nuanced optimization and is mathematically comparable to the direct preference optimization (DPO) in LLMs. Our extensive experiments on three public datasets show our method outperforms fourteen baselines in overall performance and improves the performance on tail items by a large margin. Our code is available at https://github.com/Yuhanleeee/R2Rec.

Abstract Large Language Models (LLMs) are widely applied in time series analysis. Yet, their utility in few-shot classification—a scenario with limited training data—remains unexplored. We aim to leverage the pre-trained knowledge in LLMs to overcome the data scarcity problem within multivariate time series. To this end, we propose LLMFew, an LLM-enhanced framework, to investigate the feasibility and capacity of LLMs for few-shot multivariate time series classification (MTSC). We first introduce a Patch-wise Temporal Convolution Encoder (PTCEnc) to align time series data with the textual embedding input of LLMs. Then, we fine-tune the pre-trained LLM decoder with Low-rank Adaptations (LoRA) to enable effective representation learning from time series data. Experimental results show our model consistently outperforms state-of-the-art baselines by a large margin, achieving 125.2% and 50.2% improvement in classification accuracy on Handwriting and EthanolConcentration datasets, respectively. Our results also show LLM-based methods achieve comparable performance to traditional models across various datasets in few-shot MTSC, paving the way for applying LLMs in practical scenarios where labeled data are limited. Our code is available at https://github.com/junekchen/llm-fewshot-mtsc .

Human action recognition has been explored in healthcare, sports, and entertainment, with a recent shift toward manufacturing settings for monitoring assembly tasks. Identifying assembly actions is crucial for improving human-robot collaboration and optimizing the assembly process. However, the complexity of assembly tasks poses challenges for action recognition methods, with single-modality methods struggling to capture the complex dynamics and context. We proposed the multimodal ConvLSTM-AssNet and C3D-AssNet methods, which use RGB, RGB-A, and depth data. The models are tested in single, double, and triple stream configurations, with attention mechanisms integrated to focus on relevant features. The proposed models are evaluated on the HA4M dataset. Attention-Guided C3D-AssNet is most accurate for single (RGB-A: 97.10%) and double streams (RGB-A + Depth: 98.84%), while ConvLSTM-AssNet performs best for triple streams (RGB + RGB-A + Depth: 97.30%). This research advances multimodal assembly action recognition for manufacturing applications.

Abstract Mashups are web applications that expedite software development by reusing existing resources through integrating multiple application programming interfaces (APIs). Recommending the appropriate APIs plays a critical role in assisting developers in building such web applications easily and efficiently. The proliferation of publicly available APIs on the Internet has inspired the community to adopt various models to accomplish the recommendation task. Until present, considerable efforts have been made to recommend the optimal set of APIs, delivering fruitful results and achieving varying recommendation performance. This paper presents a timely review on the topic of API recommendations for mashup creation. Specifically, we investigate and compare not only traditional data mining approaches and recommendation techniques but also more recent approaches based on network representation learning and deep learning techniques. By analyzing the merits and pitfalls of existing approaches, we pinpoint a few promising directions to address the remaining challenges in the current research. This survey provides a timely comprehensive review of the API recommendation research and could be a useful reference for relevant researchers and practitioners.

Unsupervised domain adaptation (UDA) methods have recently been explored for their use in skeleton recognition tasks. Much work along this line has been focusing on the “close-set” problems, which often deviate from reality as human actions vary in application scenarios. Thus, there remains a need to thoroughly study the “open-set” problems with UDA methods for skeleton recognition, aiming to support those models capable of self-adapting to action changes in different scenarios. To this end, we delve into the “open-set” problems from a feature alignment perspective under UDA settings in reaching domain and class alignment. Specifically, the domain-wise alignment was achieved by the maximum mean discrepancy (MMD) combined with supervision signals from the source domain, which form clear feature boundaries between the “known” and “unknown” classes. Then, the class-wise alignment was achieved by contrastive learning methods, which are distinguished from previous binary classification methods, in reaching compactness inside of “unknown” or “known” classes. Moreover, we conducted the Fourier analysis during the evaluation phases to verify the model’s robustness. To our knowledge, we are the first to apply the Fourier Heatmap in UDA methods for skeleton recognition. The heatmap visualizes the model’s sensitivity steered for interpretability. Significant performance improvements are observed on the NTU and PKU data sets when adding the domain-wise alignment module to other contrastive learning methods. Furthermore, experimental results demonstrate that our approach, termed CStrCRL-UDA, is consistent with robustness and efficiency on these two benchmark data sets.

Abstract As a significant application of machine learning in financial scenarios, loan default risk prediction aims to evaluate the client’s default probability. However, most existing deep learning solutions treat each application as an independent individual, neglecting the explicit connections among different application records. Besides, these attempts suffer from the problem of missing data and imbalanced distribution (i.e., the default records are small samples against all the applications). We believe similar records could provide some auxiliary signals, which are of critical importance to alleviate the data missing issue and facilitate data argumentation. To this end, we propose multi-view loan application graphs, dubbed MLAGs. By evaluating the similarity between the records, a loan application graph can be constructed. Furthermore, we arrange different similarity thresholds to organize various graph structures for multi-graph constructions; thus, a variety of representations can be generated via information propagation and aggregation for small sample argumentation. Consequently, the imbalanced data distribution and missing values issues can be alleviated effectively. We conduct experiments on three public datasets from real-world home credit and P2P lending platforms, which show that MGCN outperforms both conventional and deep learning models. Ablation studies also illustrated the validity of each module design.

Contrastive learning has been widely embraced for its notable success along with two augmentation methods— normal and strong augmentations—in skeleton action recognition. Existing methods gain performance largely by customizing normal augmentations while bypassing strong augmentations that riches in motion patterns. To make up for the blank, we propose a novel framework, called CStrCRL, acquiring view-invariant and discriminative features from strong augmentations by leveraging contrastive learning. Specifically, to avoid the fragility of skeleton data adversely affecting the model after applying strong augmentations, we use consistency learning to maximize the similarity between strongly and normally augmented views. Furthermore, we employ cross-view learning on strong and normal augmentations for eliminating uncertainty feature boundaries learned by the model. Moreover, we design a new backbone, termed GatedStrNet, for discriminating valid and invalid features contained in strong augmented views. Finally, extensive experiments on NTU 60/120 and PKUMMD II demonstrate that the proposed method bridges the performance gap between normal and strong augmentations on contrastive learning of skeleton recognition. Notably, with a single stream input, CStrCRL achieves accuracies of 78.93% and 84.04% on the NTU60 Xsub and Xview datasets. Our source code can be found at: https://github.com/RHu-main/CStrCRL.

introduction Free Access Share on Just AcceptedSpecial Issue on Responsible Recommender Systems Part 1 Authors: Lina Yao CSIRO's Data61 and University of New South Wales, Australia CSIRO's Data61 and University of New South Wales, Australia 0000-0002-4149-839XSearch about this author , Julian McAuley University of California, USA University of California, USASearch about this author , Xianzhi Wang University of Technology Sydney, Australia University of Technology Sydney, Australia 0000-0001-9582-3445Search about this author , Dietmar Jannach University of Klagenfurt, Austria University of Klagenfurt, Austria 0000-0002-4698-8507Search about this author Authors Info & Claims ACM Transactions on Intelligent Systems and TechnologyAccepted on April 2024https://doi.org/10.1145/3663528Online AM:15 June 2024Publication History 0citation5DownloadsMetricsTotal Citations0Total Downloads5Last 12 Months5Last 6 weeks5 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Publisher SiteeReaderPDF

Trajectory prediction is fundamental to various intelligent technologies, such as autonomous driving and robotics. The motion prediction of pedestrians and vehicles helps emergency braking, reduces collisions, and improves traffic safety. Current trajectory prediction research faces problems of complex social interactions, high dynamics and multi-modality. Especially, it still has limitations in long-time prediction. We propose Attention-aware Social Graph Transformer Networks for multi-modal trajectory prediction. We combine Graph Convolutional Networks and Transformer Networks by generating stable resolution pseudo-images from Spatio-temporal graphs through a designed stacking and interception method. Furthermore, we design the attention-aware module to handle social interaction information in scenarios involving mixed pedestrian-vehicle traffic. Thus, we maintain the advantages of the Graph and Transformer, i.e., the ability to aggregate information over an arbitrary number of neighbors and the ability to perform complex time-dependent data processing. We conduct experiments on datasets involving pedestrian, vehicle, and mixed trajectories, respectively. Our results demonstrate that our model minimizes displacement errors across various metrics and significantly reduces the likelihood of collisions. It is worth noting that our model effectively reduces the final displacement error, illustrating the ability of our model to predict for a long time.

Dynamic link prediction has become a trending research subject because of its wide applications in the web, sociology, transportation, and bioinformatics. Currently, the prevailing approach for dynamic link prediction is based on graph neural networks, in which graph representation learning is the key to perform dynamic link prediction tasks. However, there are still great challenges because the structure of graphs evolves over time. A common approach is to represent a dynamic graph as a collection of discrete snapshots, in which information over a period is aggregated through summation or averaging. This way results in some fine-grained time-related information loss, which further leads to a certain degree of performance degradation. We conjecture that such fine-grained information is vital because it implies specific behavior patterns of nodes and edges in a snapshot. To verify this conjecture, we propose a novel fine-grained behavior-aware network (BehaviorNet) for dynamic network link prediction. Specifically, BehaviorNet adapts a transformer-based graph convolution network to capture the latent structural representations of nodes by adding edge behaviors as an additional attribute of edges. GRU is applied to learn the temporal features of given snapshots of a dynamic network by utilizing node behaviors as auxiliary information. Extensive experiments are conducted on several real-world dynamic graph datasets, and the results show significant performance gains for BehaviorNet over several state-of-the-art (SOTA) discrete dynamic link prediction baselines. Ablation study validates the effectiveness of modeling fine-grained edge and node behaviors.

Temporal knowledge graph (TKG) extrapolation aims to predict future unknown events (facts) based on historical information, and has attracted considerable attention due to its great practical significance. Accurate representations (embeddings) of entities and relations form the basis of TKG extrapolation. Recent work has been devoted to improving the rationality of entity representations. However, on the one hand, ignoring relation modeling results in incomplete relation representations; therefore, some approaches aggregate only immediately adjacent entities of relations, but this can lead to the "message islands" problem of relation modeling. On the other hand, ignoring the association constraints between relations and entities can make the embeddings of both relations and entities prone to overfitting. To address the abovementioned challenges, we propose an advanced method, namely, RETIA. For the former issue, we generate twin hyperrelation subgraphs for each historical subgraph and then aggregate both the adjacent entities and relations in the hyperrelation subgraphs through a graph convolutional network (GCN). About the latter concern, we propose a twin-interact module (TIM), which provides communication channels for relation aggregation and entity aggregation during the evolution of the historical sequence. Experiments conducted on five public datasets show that RETIA has made great improvements across several evaluation metrics. Our released code is available at https://github.com/CGCL-codes/RETIA.

Improving users’ long-term experience in recommender systems (RS) has become a growing concern for recommendation platforms. Reinforcement learning (RL) is an attractive approach because it can plan and optimize long-term returns sequentially. However, directly applying RL as an online learning method in the RS setting can significantly compromise users’ satisfaction and experience. As a result, learning the recommendation policy from logged feedback collected under different policies has emerged as a promising direction. Offline learning enables the agent to utilize off-policy learning techniques. Nevertheless, several challenges need to be addressed, such as distribution shift. In this paper, we propose a novel RL method, called Supervised Off-Policy Actor-Critic (SOAC), for learning the recommendation policy from the logged feedback without exploration. The proposed SOAC addresses challenges, including distribution shift and extrapolation errors, and focuses on improving the ranking of items in a recommendation list. The experimental results demonstrate that SOAC can achieve better recommendation performance than existing supervised RL methods.

Most existing video-based posture recognition methods treat frames equally using unified or random sampling strategies, thus losing the temporal relationship information among frames. To address this problem, we propose a lightweight framework, namely RASNet, to adaptively select informative frames for recognition. Specifically, we design a video-suited exploration environment to guide the agent in learning the selection strategy. We introduce the reparametrization method to convert the discrete action space into a continuous space, making the agent robust and random. For the reward part, we design a multi-factor function to reward the agent keeping a balance between frame usage and accuracy. Extensive experiments on three large-scale datasets prove the effectiveness of RASNet, e.g., achieving 85.9% accuracy with fewer 1.15 frames than other state-of-the-art methods on Kinetics 600.

Sequential recommender systems (SRSs) aim to predict the subsequent items which may interest users via comprehensively modeling users' complex preference embedded in the sequence of user-item interactions. However, most of existing SRSs often model users' single low-level preference based on item ID information while ignoring the high-level preference revealed by item attribute information, such as item category. Furthermore, they often utilize limited sequence context information to predict the next item while overlooking richer inter-item semantic relations. To this end, in this paper, we proposed a novel hierarchical preference modeling framework to substantially model the complex low- and high-level preference dynamics for accurate sequential recommendation. Specifically, in the framework, a novel dual-transformer module and a novel dual contrastive learning scheme have been designed to discriminatively learn users' low- and high-level preference and to effectively enhance both low- and high-level preference learning respectively. In addition, a novel semantics-enhanced context embedding module has been devised to generate more informative context embedding for further improving the recommendation performance. Extensive experiments on six real-world datasets have demonstrated both the superiority of our proposed method over the state-of-the-art ones and the rationality of our design.

Graph Convolutional Networks (GCNs) are a popular type of machine learning models that use multiple layers of convolutional aggregation operations and non-linear activations to represent data. Recent studies apply GCNs to Collaborative Filtering (CF)-based recommender systems (RSs) by modeling user-item interactions as a bipartite graph and achieve superior performance. However, these models face difficulty in training with non-linear activations on large graphs. Besides, most GCN-based models could not model deeper layers due to the over-smoothing effect with the graph convolution operation. In this paper, we improve the GCN-based CF models from two aspects. First, we remove non-linearities to enhance recommendation performance, which is consistent with the theories in simple graph convolutional networks. Second, we obtain the initialization of the embedding for each node in the graph by computing the network embedding on the condensed graph, which alleviates the over smoothing problem in graph convolution aggregation operation with sparse interaction data. The proposed model is a linear model that is easy to train, scalable to large datasets, and shown to yield better efficiency and effectiveness on four real datasets.

The session-based recommendation aims to predict users' immediate next actions based on their short-term behaviors reflected by past and ongoing sessions. Graph neural networks (GNNs) recently dominated the related studies, yet their performance heavily relies on graph structures, which are often predefined, task-specific, and designed heuristically. Furthermore, existing graph-based methods either neglect implicit correlations among items or consider explicit and implicit relationships altogether in the same graphs. We propose to decouple explicit and implicit relationships among items. As such, we can capture the prior knowledge encapsulated in explicit dependencies and learned implicit correlations among items simultaneously in a flexible and more interpretable manner for effective recommendations. We design a dual graph neural network that leverages the feature representations extracted by two GNNs: a graph neural network with a single gate (SG-GNN) and an adaptive graph neural network (A-GNN). The former models explicit dependencies among items. The latter employs a self-learning strategy to capture implicit correlations among items. Our experiments on four real-world datasets show our model outperforms state-of-the-art methods by a large margin, achieving 18.46% and 70.72% improvement in [email protected], and 49.10% and 115.29% improvement in [email protected] on Diginetica and LastFM datasets.

Sequential recommendation aims to predict the next item which interests users via modeling their interest in items over time. Most of the existing works on sequential recommendation model users’ dynamic interest in specific items while overlooking users’ static interest revealed by some static attribute information of items, e.g., category, brand. Moreover, existing works often only consider the positive excitation of a user’s historical interactions on his/her next choice on candidate items while ignoring the commonly existing negative excitation, resulting in insufficiently modeling dynamic interest. The overlook of static interest and negative excitation will lead to incomplete interest modeling and thus impedes the recommendation performance. To this end, in this paper, we propose modeling both static interest and negative excitation for dynamic interest to further improve the recommendation performance. Accordingly, we design a novel Static-Dynamic Interest Learning (SDIL) framework featured with a novel Temporal Positive and Negative Excitation Modeling (TPNE) module for accurate sequential recommendation. TPNE is specially designed for comprehensively modeling dynamic interest based on temporal positive and negative excitation learning. Extensive experiments on three real-world datasets show that SDIL can effectively capture both static and dynamic interest and outperforms state-of-the-art baselines.

In light of the emergence of deep reinforcement learning (DRL) in recommender systems research and several fruitful results in recent years, this survey aims to provide a timely and comprehensive overview of recent trends of deep reinforcement learning in recommender systems. We start by motivating the application of DRL in recommender systems, followed by a taxonomy of current DRL-based recommender systems and a summary of existing methods. We discuss emerging topics, open issues, and provide our perspective on advancing the domain. The survey serves as introductory material for readers from academia and industry to the topic and identifies notable opportunities for further research.

Abstract Skin cancer, primarily resulting from the abnormal growth of skin cells, is among the most common cancer types. In recent decades, the incidence of skin cancer cases worldwide has risen significantly (one in every three newly diagnosed cancer cases is a skin cancer). Such an increase can be attributed to changes in our social and lifestyle habits coupled with devastating man-made alterations to the global ecosystem. Despite such a notable increase, diagnosis of skin cancer is still challenging, which becomes critical as its early detection is crucial for increasing the overall survival rate. This calls for advancements of innovative computer-aided systems to assist medical experts with their decision making. In this context, there has been a recent surge of interest in machine learning (ML), in particular, deep neural networks (DNNs), to provide complementary assistance to expert physicians. While DNNs have a high processing capacity far beyond that of human experts, their outputs are deterministic, i.e., providing estimates without prediction confidence. Therefore, it is of paramount importance to develop DNNs with uncertainty-awareness to provide confidence in their predictions. Monte Carlo dropout (MCD) is vastly used for uncertainty quantification; however, MCD suffers from overconfidence and being miss calibrated. In this paper, we use MCD algorithm to develop an uncertainty-aware DNN that assigns high predictive entropy to erroneous predictions and enable the model to optimize the hyper-parameters during training, which leads to more accurate uncertainty quantification. We use two synthetic (two moons and blobs) and a real dataset (skin cancer) to validate our algorithm. Our experiments on these datasets prove effectiveness of our approach in quantifying reliable uncertainty. Our method achieved 85.65 ± 0.18 prediction accuracy, 83.03 ± 0.25 uncertainty accuracy, and 1.93 ± 0.3 expected calibration error outperforming vanilla MCD and MCD with loss enhanced based on predicted entropy.

Abstract Multivariate time series classification is a critical problem in data mining with broad applications. It requires harnessing the inter-relationship of multiple variables and various ranges of temporal dependencies to assign the correct classification label of the time series. Multivariate time series may come from a wide range of sources and be used in various scenarios, bringing the classifier challenge of temporal representation learning. We propose a novel convolutional neural network architecture called Attentional Gated Res2Net for multivariate time series classification. Our model uses hierarchical residual-like connections to achieve multi-scale receptive fields and capture multi-granular temporal information. The gating mechanism enables the model to consider the relations between the feature maps extracted by receptive fields of multiple sizes for information fusion. Further, we propose two types of attention modules, channel-wise attention and block-wise attention, to better leverage the multi-granular temporal patterns. Our experimental results on 14 benchmark multivariate time-series datasets show that our model outperforms several baselines and state-of-the-art methods by a large margin. Our model outperforms the SOTA by a large margin, the classification accuracy of our model is 10.16% better than the SOTA model. Besides, we demonstrate that our model improves the performance of existing models when used as a plugin. Further, based on our experiments and analysis, we provide practical advice on applying our model to a new problem.

Product reviews on e-commerce platforms play a critical role in shaping users’ purchasing decisions. Unfortunately, online reviews sometimes can be intentionally misleading to manipulate the ecosystem. To date, existing methods to automatically detect “spam reviews” either focus on sophisticated feature engineering with traditional classification models or rely on tuning neural networks with aggregated features. In this article, we develop a novel graph-based model, namely, graph-aware deep fusion networks (GDFNs) that use information from relevant metadata (review text, features of users, and items) and relational data (network) to capture the semantic information from their complex heterogeneous interactions via graph convolutional networks (GCNs). Besides, GDFN also uses a novel fusion technique to synthesize low- and high-order interactions with propagated information across multiple review-related subgraphs. Extensive experiments on publicly available datasets show that our proposed model is effective and outperforms several strong state-of-the-art baselines.

Recent advances in reinforcement learning have inspired increasing interest in learning user modeling adaptively through dynamic interactions, e.g., in reinforcement learning based recommender systems. In most reinforcement learning applications, reward functions provide the critical guideline for optimization. However, current reinforcement learning-based methods rely on manually-defined reward functions, which cannot adapt to dynamic, noisy environments. Moreover, they generally use task-specific reward functions that sacrifice generalization ability. We propose a generative inverse reinforcement learning for user behavioral preference modeling to address the above issues. Instead of using predefined reward functions, our model can automatically learn the rewards from user's actions based on discriminative actor-critic network and Wasserstein GAN. Our model provides a general approach to characterizing and explaining underlying behavioral tendencies. Our experiments show our method outperforms state-of-the-art methods in several scenarios, namely traffic signal control, online recommender systems, and scanpath prediction.

The ability to evaluate uncertainties in evolving data streams has become equally, if not more, crucial than building a static predictor. For instance, during the pandemic, a model should consider possible uncertainties such as governmental policies, meteorological features, and vaccination schedules. Neural process families (NPFs) have recently shone a light on predicting such uncertainties by bridging Gaussian processes (GPs) and neural networks (NNs). Their abilities to output average predictions and the acceptable variances, i.e., uncertainties, made them suitable for predictions with insufficient data, such as meta-learning or few-shot learning. However, existing models have not addressed continual learning which imposes a stricter constraint on the data access. Regarding this, we introduce a member meta-continual learning with neural process (MCLNP) for uncertainty estimation. We enable two levels of uncertainty estimations: the local uncertainty on certain points and the global uncertainty p(z) that represents the function evolution in dynamic environments. To facilitate continual learning, we hypothesize that the previous knowledge can be applied to the current task, hence adopt a coreset as a memory buffer to alleviate catastrophic forgetting. The relationships between the degree of global uncertainties with the intratask diversity and model complexity are discussed. We have estimated prediction uncertainties with multiple evolving types including abrupt/gradual/recurrent shifts. The applications encompass meta-continual learning in the 1-D, 2-D datasets, and a novel spatial-temporal COVID dataset. The results show that our method outperforms the baselines on the likelihood and can rebound quickly even for heavily evolved data streams.

Tree-based models and deep neural networks are two schools of effective classification methods in machine learning. While tree-based models are robust irrespective of data domain, deep neural networks have advantages in handling high-dimensional data. Adding a differentiable neural decision forest to the neural network can generally help exploit the benefits of both models. Therefore, traditional decision trees diverge into a bagging version (i.e., random forest) and a boosting version (i.e., gradient boost decision tree). In this work, we aim to harness the advantages of both bagging and boosting by applying gradient boost to a neural decision forest. We propose a gradient boost that can learn the residual using a neural decision forest, considering the residual as a part of the final prediction. Besides, we design a structure for learning the parameters of neural decision forest and gradient boost module in contiguous steps, which is extendable to incorporate multiple gradient-boosting modules in an end-to-end manner. Our extensive experiments on several public datasets demonstrate the competitive performance and efficiency of our model against a series of baseline methods in solving various machine learning tasks.

Quality-of-service (QoS)-aware service composition aims to aggregate multiple existing services to meet users' complex functional and nonfunctional requirements that cannot be met by simple services. The accumulation of user tasks and service composition solutions makes it possible to mine empirical rules from those historical compositions to reduce the search space and thus improve the composition efficiency. Traditional empirical rule-based methods focus on mining with well-designed rules, ignoring the underlying correlations between tasks and between services. Meanwhile, infrequently used services are not valued by these methods, but these services may still be used in constructing optimal service solutions. In addition, many methods use reinforcement learning to compose services to efficiently construct service solutions, but they do not achieve the same effect as traditional metaheuristic methods. In view of the above shortcomings, considering the ability of graphs to express relationships, we first construct tasks and services as graphs and then use a graph neural network (GNN) to mine underlying correlations and predict the probability that each service will be used to construct the solution corresponding to the task. Next, based on these high-probability services, we utilize pointer network (PN)-based reinforcement learning to efficiently construct the initial service solution. The PN is often used to solve combinatorial optimization problems and is noninferior to metaheuristics for small-scale data. To increase the generalization ability of the network, we superimpose another layer on the PN. Finally, to take advantage of infrequently used services, we use the local whale optimization algorithm (WOA) to fine-tune the initial solution and obtain a high-quality solution. The experimental results show that our approach outperforms several existing methods in terms of both composition efficiency and solution quality.

Abstract Personalized decision-making can be implemented in a Federated learning (FL) framework that can collaboratively train a decision model by extracting knowledge across intelligent clients, e.g. smartphones or enterprises. FL can mitigate the data privacy risk of collaborative training since it merely collects local gradients from users without access to their data. However, FL is fragile in the presence of statistical heterogeneity that is commonly encountered in personalized decision making, e.g., non-IID data over different clients. Existing FL approaches usually update a single global model to capture the shared knowledge of all users by aggregating their gradients, regardless of the discrepancy between their data distributions. By comparison, a mixture of multiple global models could capture the heterogeneity across various clients if assigning the client to different global models (i.e., centers) in FL. To this end, we propose a novel multi-center aggregation mechanism to cluster clients using their models’ parameters. It learns multiple global models from data as the cluster centers, and simultaneously derives the optimal matching between users and centers. We then formulate it as an optimization problem that can be efficiently solved by a stochastic expectation maximization (EM) algorithm. Experiments on multiple benchmark datasets of FL show that our method outperforms several popular baseline methods. The experimental source codes are publicly available on the Github repository (GitHub repository: https://github.com/mingxuts/multi-center-fed-learning ).

Human activity recognition has become an important challenge yet to resolve while also having promising benefits in various applications for years. Existing approaches have made great progress by applying deep-learning and attention-based methods. However, the deep learning-based approaches may not fully exploit the features to resolve multimodal human activity recognition tasks. Also, the potential of attention-based methods still has not been fully explored to better extract the multimodal spatial-temporal relationship and produce robust results. In this work, we propose Multi-agent Transformer Network (MATN), a multi-agent attention-based deep learning algorithm, to address the above issues in multimodal human activity recognition. We first design a unified representation learning layer to encode the multimodal data, which preprocesses the data in a generalized and efficient way. Then we develop a multimodal spatial-temporal transformer module that applies the attention mechanism to extract the salient spatial-temporal features. Finally, we use a multi-agent training module to collaboratively select the informative modalities and predict the activity labels. We have extensively conducted experiments to evaluate MATN's performance on two public multimodal human activity recognition datasets. The results show that our model has achieved competitive performance compared to the state-of-the-art approaches, which also demonstrates scalability, effectiveness, and robustness.

Pedestrian trajectory prediction is essential for many modern applications, such as abnormal motion analysis and collision avoidance for improved traffic safety. Previous studies still face challenges in embracing high social interaction, dynamics, and multi-modality for achieving high accuracy with long-time predictions. We propose Social Graph Transformer Networks for multi-modal prediction of pedestrian trajectories, where we combine Graph Convolutional Network and Transformer Network by generating stable resolution pseudo-images from Spatio-temporal graphs through a designed stacking and interception method. Specifically, we adopt adjacency matrices to obtain Spatio-temporal features and Transformer for long-time trajectory predictions. As such, we retrain the advantages of both, i.e., the ability to aggregate information over an arbitrary number of neighbors and to conduct complex time-dependent data processing. Our experimental results show that our model reduces the final displacement error and achieves state-of-the-art in multiple metrics. The module's effectiveness is demonstrated through ablation experiments.

Multivariate time series classification is a critical problem in data mining with broad applications. We design a novel convolutional neural network architecture, Attentional Gated Res2Net, for robust multivariate time series classification. AGRes2Net uses hierarchical residual-like connections to achieve multi-scale receptive fields and to capture multi- granular temporal patterns. It further employs the gated mechanism to harness inter-relationship between feature maps. We propose two types of attention modules, namely channel-wise attention and block-wise attention, to leverage the multi-granular temporal patterns. Our experiments on six benchmark datasets demonstrate that AGRes2Net not only outperforms several baselines and state-of-the-art methods but also improves the classification accuracy of existing models when used as a plug-in.

The string similarity search is core functionality in a range of applications, including data cleaning, near-duplicate object detection, and data integration. We study the problem of threshold similarity search with the edit distance, where given a set of strings, a threshold$k$, and a query string$q$, we aim to find all strings in the set whose edit distances to$q$are no larger than$k$. Extensive studies have been proposed for the threshold similarity search problem with the edit distance. However, they suffer from a huge space consumption issue when achieving only an acceptable efficiency, especially for long strings. In this paper, we propose a simple yet small index, called minIL, to eliminate this issue. First, we adopt a minhash family to capture pivot characters and to construct sketch representations for strings. Second, we develop a multi-level inverted index to search sketches with a low space consumption. Finally, we apply a novel learned index technique on top of the index that further improves the query efficiency. Extensive experiments on real-world datasets offer insight into the performance of our method and show that it substantially reduces the index size, and is capable of outperforming the baseline approaches.

Temporal knowledge graph (TKG) reasoning has attracted significant attention. Recent approaches for modeling historical information have led to great advances. However, the problems of time variability and unseen entities have become two major obstacles preventing further development. The time variability problem means that different historical timestamps play different roles in the inference process. Furthermore, in the context of time variability, the unseen entity problem means that a query cannot obtain a predicted entity that is unseen in the scale-varying history rather than in a fixed set, thus turning from static to dynamic. In this paper, we propose a novel method named DHU-NET for addressing the time variability challenge and the dynamic unseen entity challenge derived from it. With regard to the former concern, we propose a time-distributed representation learning method based on a graph convolutional network(GCN) and a self-attention mechanism, which learns the distributed representations of facts at different historical timestamps and comprehensively pays different levels of attention to the different timestamps. With regard to the latter issue, we extract the unseen entities from a global static KG based on a copy mechanism and bring them into consideration during the final prediction step. Experiments on six benchmark datasets demonstrate the substantial improvements achieved by DHUNET in terms of multiple evaluation metrics. Our released codes are available at https://github.com/CGCL-codes/DHUNET.

Online recommendation requires handling rapidly changing user preferences. Deep reinforcement learning (DRL) is an effective means of capturing users' dynamic interest during interactions with recommender systems. Generally, it is challenging to train a DRL agent in online recommender systems because of the sparse rewards caused by the large action space (e.g., candidate item space) and comparatively fewer user interactions. Leveraging experience replay (ER) has been extensively studied to conquer the issue of sparse rewards. However, they adapt poorly to the complex environment of online recommender systems and are inefficient in learning an optimal strategy from past experience. As a step to filling this gap, we propose a novel state-aware experience replay model, in which the agent selectively discovers the most relevant and salient experiences and is guided to find the optimal policy for online recommendations. In particular, a locality-sensitive hashing method is proposed to selectively retain the most meaningful experience at scale and a prioritized reward-driven strategy is designed to replay more valuable experiences with higher chance. We formally show that the proposed method guarantees the upper and lower bound on experience replay and optimizes the space complexity, as well as empirically demonstrate our model's superiority to several existing experience replay methods over three benchmark simulation platforms.

Neural Multi-task Learning is gaining popularity as a way to learn multiple tasks jointly within a single model. While related research continues to break new ground, two major limitations still remain, including (i) poor generalization to scenarios where tasks are loosely correlated; and (ii) under-investigation on global commonality and local characteristics of tasks. Our aim is to bridge these gaps by presenting a neural multi-task learning model coined Hierarchical Task-aware Multi-headed Attention Network (HTMN). HTMN explicitly distinguishes task-specific features from task-shared features to reduce the impact caused by weak correlation between tasks. The proposed method highlights two parts: Multi-level Task-aware Experts Network that identifies task-shared global features and task-specific local features, and Hierarchical Multi-Head Attention Network that hybridizes global and local features to profile more robust and adaptive representations for each task. Afterwards, each task tower receives its hybrid task-adaptive representation to perform task-specific predictions. Extensive experiments on two real datasets show that HTMN consistently outperforms the compared methods on a variety of prediction tasks.

Abstract Recommendation is a critical tool for developing and promoting the benefits of the Internet of Things (IoT). In recent years, recommender systems have attracted considerable attention in many IoT-related fields such as smart health, smart home, smart tourism and smart marketing. However, traditional recommender system approaches fail to exploit ever-growing, dynamic and heterogeneous IoT data in building recommender systems for the IoT (RSIoT). This article aims to provide a comprehensive review of state-of-the-art RSIoT, including the related techniques, applications and a discussion on the limitations of applying recommendation systems to IoT. Finally, we propose a reference framework for comparing existing studies to guide future research and practices.

Zero-Shot Learning (ZSL) aims to transfer learned knowledge from observed classes to unseen classes via semantic correlations. A promising strategy is to learn a global-local representation that incorporates global information with extra localities (i.e., small parts/regions of inputs). However, existing methods discover localities based on explicit features without digging into the inherent properties and relationships among regions. In this work, we propose a novel Entropy-guided Reinforced Partial Convolutional Network (ERPCNet), which extracts and aggregates localities progressively based on semantic relevance and visual correlations without human-annotated regions. ERPCNet uses reinforced partial convolution and entropy guidance; it not only discovers global-cooperative localities dynamically but also converges faster for policy gradient optimization. We conduct extensive experiments to demonstrate ERPCNet’s performance through comparisons with state-of-the-art methods under ZSL and Generalized Zero-Shot Learning (GZSL) settings on four benchmark datasets. We also show that ERPCNet is time efficient and explainable through visualization analysis.

Conventional multiview clustering methods seek a view consensus through minimizing the pairwise discrepancy between the consensus and subviews. However, pairwise comparison cannot portray the interview relationship precisely if some of the subviews can be further agglomerated. To address the above challenge, we propose the agglomerative analysis to approximate the optimal consensus view, thereby describing the subview relationship within a view structure. We present an agglomerative neural network (ANN) based on constrained Laplacian rank to cluster multiview data directly without a dedicated postprocessing step (e.g., using K -means). We further extend ANN with a learnable data space to handle data of complex scenarios. Our evaluations against several state-of-the-art multiview clustering approaches on four popular data sets show the promising view-consensus analysis ability of ANN. We further demonstrate ANN's capability in analyzing complex view structures, extensibility through our case study and robustness and effectiveness of data-driven modifications.

Abstract The ability to explain why the model produced results in such a way is an important problem, especially in the medical domain. Model explainability is important for building trust by providing insight into the model prediction. However, most existing machine learning methods provide no explainability, which is worrying. For instance, in the task of automatic depression prediction, most machine learning models lead to predictions that are obscure to humans. In this work, we propose explainable Multi-Aspect Depression Detection with Hierarchical Attention Network MDHAN, for automatic detection of depressed users on social media and explain the model prediction. We have considered user posts augmented with additional features from Twitter. Specifically, we encode user posts using two levels of attention mechanisms applied at the tweet-level and word-level, calculate each tweet and words’ importance, and capture semantic sequence features from the user timelines (posts). Our hierarchical attention model is developed in such a way that it can capture patterns that leads to explainable results. Our experiments show that MDHAN outperforms several popular and robust baseline methods, demonstrating the effectiveness of combining deep learning with multi-aspect features. We also show that our model helps improve predictive performance when detecting depression in users who are posting messages publicly on social media. MDHAN achieves excellent performance and ensures adequate evidence to explain the prediction.

Zero-shot learning (ZSL) refers to the problem of learning to classify instances from novel classes (unseen) that are absent in the training set (seen). Most ZSL methods infer the correlation between visual features and attributes to train the classifier for unseen classes. They may have a strong bias towards seen classes during training. Meta-learning has been introduced to mitigate the basis, but meta-ZSL methods are inapplicable when tasks used for training are sampled from diverse distributions. In this regard, we propose a novel Task-aligned Generative Meta-learning model for Zero-shot learning (TGMZ), aiming to mitigate the potentially biased training and to enable meta-ZSL to accommodate real-world datasets that contain diverse distributions. Specifically, TGMZ incorporates an attribute-conditioned task-wise distribution alignment network that projects tasks into a unified distribution to deliver an unbiased model. Our experiments show TGMZ achieves a relative improvement of 2.1%, 3.0%, 2.5%, and 7.6% over state-of-the-art algorithms on AWA1, AWA2, CUB, and aPY datasets, respectively. Overall, TGMZ outperforms competitors by 3.6% in the generalized zero-shot learning (GZSL) setting and 7.9% in our proposed fusion-ZSL setting.

Deep reinforcement learning enables an agent to capture users' interest through dynamic interactions with the environment. It uses a reward function to learn user's interest and to control the learning process, attracting great interest in recommendation research. However, most reward functions are manually designed; they are either too unrealistic or imprecise to reflect the variety, dimensionality, and non-linearity of the recommendation problem. This impedes the agent from learning an optimal policy in highly dynamic online recommendation scenarios. To address the above issue, we propose a generative inverse reinforcement learning approach that avoids the need of defining an elaborative reward function. In particular, we model the recommendation problem as an automatic policy learning problem. We first generate policies based on observed users' preferences and then evaluate the learned policy by a measurement based on a discriminative actor-critic network. We conduct experiments on an online platform, VirtualTB, and demonstrate the feasibility and effectiveness of our proposed approach via comparisons with several state-of-the-art methods.

The ability to deal with uncertainty in machine learning models has become equally, if not more, crucial to their predictive ability itself. For instance, during the pandemic, governmental policies and personal decisions are constantly made around uncertainties. Targeting this, Neural Process Families (NPFs) have recently shone a light on prediction with uncertainties by bridging Gaussian processes and neural networks. Latent neural process, a member of NPF, is believed to be capable of modelling the uncertainty on certain points (local uncertainty) as well as the general function priors (global uncertainties). Nonetheless, some critical questions remain unresolved, such as a formal definition of global uncertainties, the causality behind global uncertainties, and the manipulation of global uncertainties for generative models. Regarding this, we build a member GloBal Convolutional Neural Process(GBCoNP) that achieves the SOTA log-likelihood in latent NPFs. It designs a global uncertainty representation p(z), which is an aggregation on a discretized input space. The causal effect between the degree of global uncertainty and the intra-task diversity is discussed. The learnt prior is analyzed on a variety of scenarios, including 1D, 2D, and a newly proposed spatial-temporal COVID dataset. Our manipulation of the global uncertainty not only achieves generating the desired samples to tackle few-shot learning, but also enables the probability evaluation on the functional priors.

Deep learning frameworks generally require sufficient training data to generalize well while fail to adapt on small or few-shot datasets. Meta-learning offers an effective means of tackling few-shot scenarios and has drawn increasing attention in recent years. Meta-optimization aims to learn a shared set of parameters across tasks for meta-learning while facing challenges in determining whether an initialization condition can be generalized to tasks with diverse distributions. In this regard, we propose a meta-gradient boosting framework that can fit diverse distributions based on a base learner (which learns shared information across tasks) and a series of gradient-boosted modules (which capture task-specific information). We evaluate the model on several few-shot learning benchmarks and demonstrate the effectiveness of our model in modulating task-specific meta-learned priors and handling diverse distributions.

Understanding customer behaviour (UCB) sequences with the multi-dimensional and temporal of the data is necessary for any competitive and global business aiming to provide interesting insights and to improve business strategies. While existing researchers have applied various data analytics approaches to understand and analyze behaviors of customer, they often failed to allow the analysts including business management, product marketing and development, and decision making, etc. Achieving these goals in collaboration with domain experts, we conducted a design study contributes to address a known problem with a novel solution and to provide data-driven visual decision support in collective policy data. We determine core study demands and then use a Visual Interactive System for Understanding Customer Behavior, named UCBVis that enables decision makers to gain detail insights into customer activities. In this study, we present customer behaviour pattern of multidimensional relationship through the visualisation system based on interweaving the pattern mining and querying with a designed encoding scheme. We use a large number of customer claim records and present visual outcomes to facilitate the exploration of customer behavior. Furthermore, we provide a concise set of insights and challenges associated with the use of UCBVis in the life insurance industry. We show the robustness of UCBVis through a user study with five participants shows that UCBVis is perceived to be more useful and provides actionable insights.

Mashups are a dominant approach for building data-centric applications, especially mobile applications, in recent years. Since mashups are predominantly based on public data sources and existing APIs, it requires no sophisticated programming knowledge of people to develop mashup applications. The recent prevalence of open APIs and open data sources in the Big Data era has provided new opportunities for mashup development, but at the same time increase the difficulty of selecting the right services for a given mashup task. The API recommendation for mashup differs from traditional service recommendation tasks in lacking the specific QoS information and formal semantic specification of the APIs, which limits the adoption of many existing methods. Although there are a significant number of service recommendation approaches, most of them focus on improving the recommendation accuracy and work pays attention to the diversity of the recommendation results. Another challenge comes from the existence of both explicit and implicit correlations among the different APIs, which are generally neglected by existing recommendation methods. In this paper, we address the above deficiencies of existing approaches by exploring API recommendation for mashups in the reusable composition context, with the goal of helping developers identify the most appropriate APIs for their composition tasks. In particular, we propose a probabilistic matrix factorization approach with implicit correlation regularization to solve the recommendation problem and enhance the recommendation diversity. We conjecture that the co-invocation of APIs in real-world mashups is driven by both the explicit textual similarity and implicit correlations of APIs such as the similarity or the complementary relationship of APIs. We develop a latent variable model to uncover the latent correlations between APIs by analyzing their co-invocation patterns. We further explore the relationships of topics/categories to the proposed approach. We demonstrate the effectiveness of our approach by conducting extensive experiments on a real dataset crawled from ProgrammableWeb.

Zero-shot object detection (ZSD) has received considerable attention from the community of computer vision in recent years. It aims to simultaneously locate and categorize previously unseen objects during inference. One crucial problem of ZSD is how to accurately predict the label of each object proposal, i.e. categorizing object proposals, when conducting ZSD for unseen categories. Previous ZSD models generally relied on learning an embedding from visual space to semantic space or learning a joint embedding between semantic description and visual representation. As the features in the learned semantic space or the joint projected space tend to suffer from the hubness problem, namely the feature vectors are likely embedded to an area of incorrect labels, and thus it will lead to lower detection precision. In this paper, instead, we propose to learn a deep embedding from the semantic space to the visual space, which enables to well alleviate the hubness problem, because, compared with semantic space or joint embedding space, the distribution in visual space has smaller variance. After learning a deep embedding model, we perform $k$ nearest neighbor search in the visual space of unseen categories to determine the category of each semantic description. Extensive experiments on two public datasets show that our approach significantly outperforms the existing methods.

Interactive recommendation aims to learn from dynamic interactions between items and users to achieve responsiveness and accuracy. Reinforcement learning is inherently advantageous for coping with dynamic environments and thus has attracted increasing attention in interactive recommendation research. Inspired by knowledge-aware recommendation, we proposed Knowledge-Guided deep Reinforcement learning (KGRL) to harness the advantages of both reinforcement learning and knowledge graphs for interactive recommendation. This model is implemented upon the actor-critic network framework. It maintains a local knowledge network to guide decision-making and employs the attention mechanism to capture long-term semantics between items. We have conducted comprehensive experiments in a simulated online environment with six public real-world datasets and demonstrated the superiority of our model over several state-of-the-art methods.

Most current studies on survey analysis and risk tolerance modelling lack professional knowledge and domain-specific models. Given the effectiveness of generative adversarial learning in cross-domain information, we design an Asymmetric cross-Domain Generative Adversarial Network (ADGAN) for domain scale inequality. ADGAN utilizes the information-sufficient domain to provide extra information to improve the representation learning on the information-insufficient domain via domain alignment. We provide data analysis and user model on two data sources: Consumer Consumption Information and Survey Information. We further test ADGAN on a real-world dataset with view embedding structures and show ADGAN can better deal with the class imbalance and unqualified data space than state-of-the-art, demonstrating the effectiveness of leveraging asymmetrical domain information.

Zero-shot object detection aims to detect and recognize objects unobserved in training samples from images. Previous studies generally utilized concept names or textual descriptions to build relationships between seen and unseen classes. However, these works rarely exploited the valuable information in textual descriptions for optimizing the network. Actually, textual descriptions contain much valuable information related to categories. Exploiting this information can help training the network and improve the detection performance. Besides, textual descriptions usually contain the names of objects that need to be detected. By using this character, we can narrow the scope of candidate unseen categories, thus can improve the detection accuracy. In this regard, we propose a novel framework that incorporates both images and their text descriptions for zero-shot object detection. In particular, we employ text convolutional neural network (CNN) and Faster R-CNN to extract text features and image features respectively, and combine them to optimize the regions that contain objects in images and to classify those newly detected objects simultaneously. Besides, we try extracting potential object labels directly from textual descriptions and introducing online hard example mining (OHEM) to assist with object classification and network optimization. Our extensive experiments on two public datasets demonstrate the superior performance of our approach to state-of-the-art methods.

It has been a significant challenge to portray intraclass disparity precisely in the area of activity recognition, as it requires a robust representation of the correlation between subject-specific variation for each activity class. In this work, we propose a novel end-to-end knowledge directed adversarial learning framework, which portrays the class-conditioned intraclass disparity using two competitive encoding distributions and learns the purified latent codes by denoising learned disparity. Furthermore, the domain knowledge is incorporated in an unsupervised manner to guide the optimization and further boosts the performance. The experiments on four HAR benchmark datasets demonstrate the robustness and generalization of our proposed methods over a set of state-of-the-art. We further prove the effectiveness of automatic domain knowledge incorporation in performance enhancement.

Modeling complex spatial and temporal correlations in the correlated time series data is indispensable for understanding the traffic dynamics and predicting the future status of an evolving traffic system. Recent works focus on designing complicated graph neural network architectures to capture shared patterns with the help of pre-defined graphs. In this paper, we argue that learning node-specific patterns is essential for traffic forecasting while pre-defined graph is avoidable. To this end, we propose two adaptive modules for enhancing Graph Convolutional Network (GCN) with new capabilities: 1) a Node Adaptive Parameter Learning (NAPL) module to capture node-specific patterns; 2) a Data Adaptive Graph Generation (DAGG) module to infer the inter-dependencies among different traffic series automatically. We further propose an Adaptive Graph Convolutional Recurrent Network (AGCRN) to capture fine-grained spatial and temporal correlations in traffic series automatically based on the two modules and recurrent networks. Our experiments on two real-world traffic datasets show AGCRN outperforms state-of-the-art by a significant margin without pre-defined graphs about spatial connections.

Human Activity Recognition (HAR) plays an irreplaceable role in various applications such as security, gaming, and assisted living. Recent studies introduce deep learning to mitigate the manual feature extraction (i.e., data representation) efforts and achieve high accuracy. However, there are still challenges in learning accurate representations for sensory data due to the weakness of representation modules and the subject variances. We propose a scheme called Distance-based HAR from Ensembled spatial-temporal Representations (DHARER) to address above challenges. The idea behind DHARER is straightforward—the same activities should have similar representations. We first learn representations of the input sensory segments and latent prototype representations of each class, using a Convolution Neural Network (CNN)-based dual-stream representation module; then the learned representations are projected to activity types by measuring their similarity to the learned prototypes. We have conducted extensive experiments under a strict subject-independent setting on three large-scale datasets to evaluate the proposed scheme, and our experimental results demonstrate superior performance of DHARER to several state-of-the-art methods.

Adversarial attacks pose significant challenges for detecting adversarial attacks at an early stage. We propose attack-agnostic detection on reinforcement learning-based interactive recommendation systems. We first craft adversarial examples to show their diverse distributions and then augment recommendation systems by detecting potential attacks with a deep learning-based classifier based on the crafted data. Finally, we study the attack strength and frequency of adversarial examples and evaluate our model on standard datasets with multiple crafting methods. Our extensive experiments show that most adversarial attacks are effective, and both attack strength and attack frequency impact the attack performance. The strategically-timed attack achieves comparative attack performance with only 1/3 to 1/2 attack frequency. Besides, our black-box detector trained with one crafting method has the generalization ability over several crafting methods.

Compared with individual recommendation, recommending services to a group of users is more complicated because of various users' preference should be considered and introduces new challenging such as fairness, which has never been well studied in current works. In this paper, we propose a novel recommendation scheme called PFGR, which combines a probabilistic model with coalition game strategy, to ensure the accuracy and fairness between groups of users. Given a group of users and a set of services, PFGR models a generative process for service selection in light of several observations: 1) each group is related with several topics; 2) users' decisions on the service selection depends on their expertise, the opinions of members they are familiar with, and group influence; 3) each group contains active users and inactive user, whose activeness contributes to the existence of group. PFGR first estimates the preference of each user on a candidate service via combining user's expertise, inherent connection, and group influence. Then, it determines a group's decision on a service by aggregating the preference of group members using adaptive weights. Finally, PFGR considers users' activeness and employs a strategy based on coalition game to produce a ranked list which is fair to each group member as much as possible. Experimental results on three real-world datasets validate that PFGR can achieve higher Hit Rate and Average Reciprocal Hit Rank than state-of-the-art approaches, which indicates that PFGR attains both the precision and fairness of recommendation.

Truth discovery has been widely studied in recent years as a fundamental means for resolving the conflicts in multi-source data. Although many truth discovery methods have been proposed based on different considerations and intuitions, investigations show that no single method consistently outperforms the others. To select the right truth discovery method for a specific application scenario, it becomes essential to evaluate and compare the performance of different methods. A drawback of current research efforts is that they commonly assume the availability of certain ground truth for the evaluation of methods. However, the ground truth may be very limited or even impossible to obtain, rendering the evaluation biased. In this article, we present CompTruthHyp , a generic approach for comparing the performance of truth discovery methods without using ground truth. In particular, our approach calculates the probability of observations in a dataset based on the output of different methods. The probability is then ranked to reflect the performance of these methods. We review and compare 12 representative truth discovery methods and consider both single-valued and multi-valued objects. The empirical studies on both real-world and synthetic datasets demonstrate the effectiveness of our approach for comparing truth discovery methods.

Recent years have witnessed the success of deep learning methods in human activity recognition (HAR). The longstanding shortage of labeled activity data inherently calls for a plethora of semisupervised learning methods, and one of the most challenging and common issues with semisupervised learning is the imbalanced distribution of labeled data over classes. Although the problem has long existed in broad real-world HAR applications, it is rarely explored in the literature. In this paper, we propose a semisupervised deep model for imbalanced activity recognition from multimodal wearable sensory data. We aim to address not only the challenges of multimodal sensor data (e.g., interperson variability and interclass similarity) but also the limited labeled data and class-imbalance issues simultaneously. In particular, we propose a pattern-balanced semisupervised framework to extract and preserve diverse latent patterns of activities. Furthermore, we exploit the independence of multi-modalities of sensory data and attentively identify salient regions that are indicative of human activities from inputs by our recurrent convolutional attention networks. Our experimental results demonstrate that the proposed model achieves a competitive performance compared to a multitude of state-of-the-art methods, both semisupervised and supervised ones, with 10% labeled training data. The results also show the robustness of our method over imbalanced, small training data sets.

Object detection is important in real-world applications. Existing methods mainly focus on object detection with sufficient labelled training data or zero-shot object detection with only concept names. In this paper, we address the challenging problem of zero-shot object detection with natural language description, which aims to simultaneously detect and recognize novel concept instances with textual descriptions. We propose a novel deep learning framework to jointly learn visual units, visual-unit attention and word-level attention, which are combined to achieve word-proposal affinity by an element-wise multiplication. To the best of our knowledge, this is the first work on zero-shot object detection with textual descriptions. Since there is no directly related work in the literature, we investigate plausible solutions based on existing zero-shot object detection for a fair comparison. We conduct extensive experiments on three challenging benchmark datasets. The extensive experimental results confirm the superiority of the proposed model.

Online ride-sharing platforms have become a critical part of the urban transportation system. Accurately recommending hotspots to drivers in such platforms is essential to help drivers find passengers and improve users' experience, which calls for efficient passenger demand prediction strategy. However, predicting multi-step passenger demand is challenging due to its high dynamicity, complex dependencies along spatial and temporal dimensions, and sensitivity to external factors (meteorological data and time meta). We propose an end-to-end deep learning framework to address the above problems. Our model comprises three components in pipeline: 1) a cascade graph convolutional recurrent neural network to accurately extract the spatial-temporal correlations within citywide historical passenger demand data; 2) two multi-layer LSTM networks to represent the external meteorological data and time meta, respectively; 3) an encoder-decoder module to fuse the above two parts and decode the representation to predict over multi-steps into the future. The experimental results on three real-world datasets demonstrate that our model can achieve accurate prediction and outperform the most discriminative state-of-the-art methods.

Electroencephalography (EEG) signals reflect and measure activities in certain brain areas. Its zero clinical risk and easy-to-use features make it a good choice of providing insights into the cognitive process. However, effective analysis of time-varying EEG signals remains challenging. First, EEG signal processing and feature engineering are time-consuming and highly rely on expert knowledge, and most existing studies focus on domain-specific classification algorithms, which may not apply to other domains. Second, EEG signals usually have low signal-to-noise ratios and are more chaotic than other sensor signals. In this regard, we propose a generic EEG-based cognitive activity recognition framework that can adaptively support a wide range of cognitive applications to address the above issues. The framework uses a reinforced selective attention model to choose the characteristic information among raw EEG signals automatically. It employs a convolutional mapping operation to dynamically transform the selected information into a feature space to uncover the implicit spatial dependency of EEG sample distribution. We demonstrate the effectiveness of the framework under three representative scenarios: intention recognition with motor imagery EEG, person identification, and neurological diagnosis, and further evaluate it on three widely used public datasets. The experimental results show our framework outperforms multiple state-of-the-art baselines and achieves competitive accuracy on all the datasets while achieving low latency and high resilience in handling complex EEG signals across various domains. The results confirm the suitability of the proposed generic approach for a range of problems in the realm of brain-computer Interface applications.

Multi-step passenger demand forecasting is a crucial task in on-demand vehicle sharing services. However, predicting passenger demand is generally challenging due to the nonlinear and dynamic spatial-temporal dependencies. In this work, we propose to model multi-step citywide passenger demand prediction based on a graph and use a hierarchical graph convolutional structure to capture both spatial and temporal correlations simultaneously. Our model consists of three parts: 1) a long-term encoder to encode historical passenger demands; 2) a short-term encoder to derive the next-step prediction for generating multi-step prediction; 3) an attention-based output module to model the dynamic temporal and channel-wise information. Experiments on three real-world datasets show that our model consistently outperforms many baseline methods and state-of-the-art models.

The prevalence of online social media facilitates massive knowledge acquisition and sharing throughout the Web. Meanwhile, it inevitably poses the risk of generating and disseminating false information by both benign and malicious users. Despite there has been considerable research on false information detection from both the opinion-based and fact-based perspectives, they mostly focus on tailored solutions for a particular domain and carry out limited work on leveraging multi-faceted clues such as textual cues, behavioral trails, and relational connection. We propose a novel dual-stream attentive random forest that is capable of selecting clues of discriminative information from individuals, collective information (e.g., texts), and correlations of entities (e.g., social interactions) adaptively. In particular, we use an interpretive attention model for learning textual contents. The model treats the important and unimportant content differently when constructing the textual representation and employs a multilayer perceptron to capture the hidden complex relationships among features of side information. We further propose a unified framework for leveraging the above clues, where we use attentive forests to provide probabilistic distribution as predictions over the two learned representations, which are then leveraged to make a better estimation. We conduct extensive experiments on three real-world benchmark datasets for fake news and fake review detection. The results show our approach outperforms multiple baselines in the accuracy of detecting false information.

Service selection is crucial to service composition in determining the composite Quality of Service (QoS). The proliferation of composable services on the Internet and the practical need for timely delivering optimized composite solutions motivate the adoption of population-based algorithms for QoS-aware service selection. However, existing population-based algorithms are generally complicated to use, and often used as a general approach to solving different optimization problems. We propose to develop specialized algorithms for QoS-aware service selection, based on the artificial bee colony algorithm (ABC). ABC is a new and simpler implementation of swarm intelligence, which has proven to be successful in solving many real-world problems, especially the numerical optimization problems. We develop an approximate approach for the neighborhood search of ABC, which enables effective local search in the discrete space of service selection in a way that is analogical to the search in a continuous space. We present three algorithms based on the approach. All the three algorithms are designed to improve the performance and meanwhile preserve the simplicity of ABC. Each algorithm applies a different technique to leverage the unique characteristics of the service selection problem. Experimental results show higher accuracy and convergence speed of the proposed algorithms over the state of the art algorithms.

Multimodal features play a key role in wearable sensor based human activity recognition (HAR). Selecting the most salient features adaptively is a promising way to maximize the effectiveness of multimodal sensor data. In this regard, we propose a “collect fully and select wisely” principle as well as an interpretable parallel recurrent model with convolutional attentions to improve the recognition performance. We first collect modality features and the relations between each pair of features to generate activity frames, and then introduce an attention mechanism to select the most prominent regions from activity frames precisely. The selected frames not only maximize the utilization of valid features but also reduce the number of features to be computed effectively. We further analyze the accuracy and interpretability of the proposed model based on extensive experiments. The results show that our model achieves competitive performance on two benchmarked datasets and works well in real life scenarios.

With the widespread adoption of Internet of Things (IoT), billions of everyday objects are being connected to the Internet. Effective management of these devices to support reliable, secure and high quality applications becomes challenging due to the scale. As one of the key cornerstones of IoT device management, automatic cross-device classification aims to identify the semantic type of a device by analyzing its network traffic. It has the potential to underpin a broad range of novel features such as enhanced security (by imposing the appropriate rules for constraining the communications of certain types of devices) or context-awareness (by the utilization and interoperability of IoT devices and their high-level semantics) of IoT applications. We propose an automatic IoT device classification method to identify new and unseen devices. The method uses the rich information carried by the traffic flows of IoT networks to characterize the attributes of various devices. We first specify a set of discriminating features from raw network traffic flows, and then propose a LSTM-CNN cascade model to automatically identify the semantic type of a device. Our experimental results using a real-world IoT dataset demonstrate that our proposed method is capable of delivering satisfactory performance. We also present interesting insights and discuss the potential extensions and applications.

The global urbanization imposes unprecedented pressure on urban infrastructure and public resources. The population explosion has made it challenging to satisfy the daily needs of urban residents. 'Smart City' is a solution that utilizes different types of data collection sensors to help manage assets and resources intelligently and more efficiently. Under the Smart City umbrella, the primary research initiative in improving the efficiency of car-hailing services is to predict the citywide passenger demand to address the imbalance between the demand and supply. However, predicting the passenger demand requires analysis on various data such as historical passenger demand, crowd outflow, and weather information, and it remains challenging to discover the latent relationships among these data. To address this challenge, we propose to improve the passenger demand prediction via learning the salient spatial-temporal dynamics within a reinforcement learning framework. Our model employs an information selection mechanism to focus on the most distinctive data in historical observations. This mechanism can automatically adjust the information zone according to the prediction performance to find the optimal choice. It also ensures the prediction model to take full advantage of the available data by introducing the positive and excluding the negative correlations. We have conducted experiments on a large-scale real-world dataset that covers 1.5 million people in a major city in China. The results show our model outperforms state-of-the-art and a series of baselines by a large margin.

Understanding and recognizing human activities is a fundamental research topic for a wide range of important applications such as fall detection and remote health monitoring and intervention. Despite active research in human activity recognition over the past years, existing approaches based on computer vision or wearable sensor technologies present several significant issues such as privacy (e.g., using video camera to monitor the elderly at home) and practicality (e.g., not possible for an older person with dementia to remember wearing devices). In this paper, we present a low-cost, unobtrusive, and robust system that supports independent living of older people. The system interprets what a person is doing by deciphering signal fluctuations using radio-frequency identification (RFID) technology and machine learning algorithms. To deal with noisy, streaming, and unstable RFID signals, we develop a compressive sensing, dictionary-based approach that can learn a set of compact and informative dictionaries of activities using an unsupervised subspace decomposition. In particular, we devise a number of approaches to explore the properties of sparse coefficients of the learned dictionaries for fully utilizing the embodied discriminative information on the activity recognition task. Our approach achieves efficient and robust activity recognition via a more compact and robust representation of activities. Extensive experiments conducted in a real-life residential environment demonstrate that our proposed system offers a good overall performance and shows the promising practical potential to underpin the applications for the independent living of the elderly.

Point-of-Interest (POI) recommendation is a new type of recommendation task that comes along with the prevalence of location-based social networks and services in recent years. Compared with traditional recommendation tasks, POI recommendation focuses more on making personalized and context-aware recommendations to improve user experience. Traditionally, the most commonly used contextual information includes geographical and social context information. However, the increasing availability of check-in data makes it possible to design more effective location recommendation applications by modeling and integrating comprehensive types of contextual information, especially the temporal information. In this article, we propose a collaborative filtering method based on Tensor Factorization, a generalization of the Matrix Factorization approach, to model the multi-dimensional contextual information. Tensor Factorization naturally extends Matrix Factorization by increasing the dimensionality of concerns, within which the three-dimensional model is the one most popularly used. Our method exploits a high-order tensor to fuse heterogeneous contextual information about users’ check-ins instead of the traditional two-dimensional user-location matrix. The factorization of this tensor leads to a more compact model of the data that is naturally suitable for integrating contextual information to make POI recommendations. Based on the model, we further improve the recommendation accuracy by utilizing the internal relations within users and locations to regularize the latent factors. Experimental results on a large real-world dataset demonstrate the effectiveness of our approach.

Question answering (Q&A) communities have gained momentum recently as an effective means of knowledge sharing over the crowds, where many users are experts in the real-world and can make quality contributions in certain domains or technologies. Although the massive user-generated Q&A data present a valuable source of human knowledge, a related challenging issue is how to find those expert users effectively. In this paper, we propose a framework for finding such experts in a collaborative network. Accredited with recent works on distributed word representations, we are able to summarize text chunks from the semantics perspective and infer knowledge domains by clustering pre-trained word vectors. In particular, we exploit a graph-based clustering method for knowledge domain extraction and discern the shared latent factors using matrix factorization techniques. The proposed clustering method features requiring no post-processing of clustering indicators and the matrix factorization method is combined with the semantic similarity of the historical answers to conduct expertise ranking of users given a query. We use Stack Overflow, a website with a large group of users and a large number of posts on topics related to computer programming, to evaluate the proposed approach and conduct extensively experiments to show the effectiveness of our approach.

The Internet of Things (IoT) changes many sectors of our lives. In the healthcare domain, IoT presents as mobile medical applications over various sensors that update healthcare professionals on patients' health information. However, IoT-based healthcare systems also face major challenges in protecting patients' privacy via an effective access control system. This paper presents an ambient home solution framework for privacy-preserving monitoring of patients' health status. We focus on two major points: 1) how to use the data collected from ambient and biometric sensors, to perform the high-level task of activity recognition, and 2) how to secure the collected healthcare data via effective access control. We achieve multi-level access control by using Public Key Infrastructure (PKI) for authentication and Attribute-Based Access Control (ABAC) for authorisation. Our access control system regulates access to healthcare data by classification over healthcare professionals and data. Our system provides guidelines to define data classes and healthcare professional groups and specifies security policies to control access to the data classes. The system is flexible and can incorporate more policy rules, professionals, and data classes.

An electroencephalography (EEG) based brain activity recognition is a fundamental field of study for a number of significant applications such as intention prediction, appliance control, and neurological disease diagnosis in smart home and smart healthcare domains. Existing techniques mostly focus on binary brain activity recognition for a single person, which limits their deployment in wider and complex practical scenarios. Therefore, multi-person and multi-class brain activity recognition has obtained popularity recently. Another challenge faced by brain activity recognition is the low recognition accuracy due to the massive noises and the low signal-to-noise ratio in EEG signals. Moreover, the feature engineering in EEG processing is time-consuming and highly relies on the expert experience. In this paper, we attempt to solve the above challenges by proposing an approach which has better EEG interpretation ability via raw Electroencephalography (EEG) signal analysis for multi-person and multi-class brain activity recognition. Specifically, we analyze inter-class and inter-person EEG signal characteristics, based on which to capture the discrepancy of inter-class EEG data. Then, we adopt an Autoencoder layer to automatically refine the raw EEG signals by eliminating various artifacts. We evaluate our approach on both a public and a local EEG datasets and conduct extensive experiments to explore the effect of several factors (such as normalization methods, training data size, and Autoencoder hidden neuron size) on the recognition results. The experimental results show that our approach achieves a high accuracy comparing to competitive state-of-the-art methods, indicating its potential in promoting future research on multi-person EEG recognition.

Data veracity is a grand challenge for various tasks on the Web. Since the web data sources are inherently unreliable and may provide conflicting information about the same real-world entities, truth discovery is emerging as a countermeasure of resolving the conflicts by discovering the truth, which conforms to the reality, from the multi-source data. A major challenge related to truth discovery is that different data items may have varying numbers of true values (or multi-truth), which counters the assumption of existing truth discovery methods that each data item should have exactly one true value. In this paper, we address this challenge by exploiting and leveraging the implications from multi-source data. In particular, we exploit three types of implications, namely the implicit negative claims, the distribution of positive/negative claims, and the co-occurrence of values in sources' claims, to facilitate multi-truth discovery. We propose a probabilistic approach with improvement measures that incorporate the three implications in all stages of truth discovery process. In particular, incorporating the negative claims enables multi-truth discovery, considering the distribution of positive/negative claims relieves truth discovery from the impact of sources' behavioral features in the specific datasets, and considering values' co-occurrence relationship compensates the information lost from evaluating each value in the same claims individually. Experimental results on three real-world datasets demonstrate the effectiveness of our approach.

Truth discovery is the problem of detecting true values from the conflicting data provided by multiple sources on the same data items. Since sources' reliability is unknown a priori, a truth discovery method usually estimates sources' reliability along with the truth discovery process. A major limitation of existing truth discovery methods is that they commonly assume exactly one true value on each data item and therefore cannot deal with the more general case that a data item may have multiple true values (or multi-truth). Since the number of true values may vary from data item to data item, this requires truth discovery methods being able to detect varying numbers of truth values from the multi-source data. In this paper, we propose a multi-truth discovery approach, which addresses the above challenges by providing a generic framework for enhancing existing truth discovery methods. In particular, we redeem the numbers of true values as an important clue for facilitating multi-truth discovery. We present the procedure and components of our approach, and propose three models, namely the byproduct model, the joint model, and the synthesis model to implement our approach. We further propose two extensions to enhance our approach, by leveraging the implications of similar numerical values and values' co-occurrence information in sources' claims to improve the truth discovery accuracy. Experimental studies on real-world datasets demonstrate the effectiveness of our approach.

With the unprecedented and dramatic development of Web services in recent years, designing novel approaches for efficient Web service prediction has become of paramount importance. Quality of Service (QoS) plays a critical role in Web service recommendation. However determining QoS values of Web services is still a challenging task. For example, some QoS properties (e.g., response time, throughput) may hold different values for different users. In this paper, we describe how to develop a novel approach, PLMwsp, based on a probabilistic latent model, to predict effectively the QoS values of Web services. A Web service prediction has been developed, and experiments have been conducted to show the efficacy of our approach.

In order to suppress the zero sequence circulating currents (ZSCCs) between parallel operating three level voltage source inverters with common AC and DC buses, a common mode voltage reduction PWM (CMVR-PWM) technique and neural point potentials (NPPs) control based method is proposed in this paper. An equivalent model of ZSCC is developed, and two excitation sources of ZSCC, differences of common mode voltages (CMVs) and NPPs between paralleled inverters, are analyzed. A CMVR-PWM method is investigated to reduce the CMV, and a simple electric circuit is adopted to control the NPPs. With this two strategies, ZSCCs between parallel inverters can be eliminated effectively. This strategy has the advantage of without carrier synchronization and can be utilized to parallel operating inverters with different types of filter. Simulation result validate the proposed ZSCC elimination schemes.

Many real-world applications rely on multiple data sources to provide information on their interested items. Due to the noises and uncertainty in data, given a specific item, the information from different sources may conflict. To make reliable decisions based on these data, it is important to identify the trustworthy information by resolving these conflicts, i.e., the truth discovery problem. Current solutions to this problem detect the veracity of each value jointly with the reliability of each source for each data item. In this way, the efficiency of truth discovery is strictly confined by the problem scale, which in turn limits truth discovery algorithms from being applicable on a large scale. To address this issue, we propose an approximate truth discovery approach, which divides sources and values into groups according to a user-specified approximation criterion. The groups are then used for efficient inter-value influence computation to improve the accuracy. Our approach is applicable to most existing truth discovery algorithms. Experiments on real-world datasets show that our approach improves the efficiency compared to existing algorithms while achieving similar or even better accuracy. The scalability is further demonstrated by experiments on large synthetic datasets.

Truth-finding is the fundamental technique for corroborating reports from multiple sources in both data integration and collective intelligent applications. Traditional truth-finding methods assume a single true value for each data item and therefore cannot deal will multiple true values (i.e., the multi-truth-finding problem). So far, the existing approaches handle the multi-truth-finding problem in the same way as the single-truth-finding problems. Unfortunately, the multi-truth-finding problem has its unique features, such as the involvement of sets of values in claims, different implications of inter-value mutual exclusion, and larger source profiles. Considering these features could provide new opportunities for obtaining more accurate truth-finding results. Based on this insight, we propose an integrated Bayesian approach to the multi-truth-finding problem, by taking these features into account. To improve the truth-finding efficiency, we reformulate the multi-truth-finding problem model based on the mappings between sources and (sets of) values. New mutual exclusive relations are defined to reflect the possible co-existence of multiple true values. A finer-grained copy detection method is also proposed to deal with sources with large profiles. The experimental results on three real-world datasets show the effectiveness of our approach.

In this paper, we explore service recommendation and selection in the reusable composition context. The goal is to aid developers finding the most appropriate services in their composition tasks. We specifically focus on mashups, a domain that increasingly targets people without sophisticated programming knowledge. We propose a probabilistic matrix factorization approach with implicit correlation regularization to solve this problem. In particular, we advocate that the co-invocation of services in mashups is driven by both explicit textual similarity and implicit correlation of services, and therefore develop a latent variable model to uncover the latent connections between services by analyzing their co-invocation patterns. We crawled a real dataset from Programmable Web, and extensively evaluated the effectiveness of our proposed approach.

Point-of-Interest (POI) recommendation is a new type of recommendation task that comes along with the prevalence of location-based social networks in recent years. Compared with traditional tasks, it focuses more on personalized, context-aware recommendation results to provide better user experience. To address this new challenge, we propose a Collaborative Filtering method based on Non-negative Tensor Factorization, a generalization of the Matrix Factorization approach that exploits a high-order tensor instead of traditional User-Location matrix to model multi-dimensional contextual information. The factorization of this tensor leads to a compact model of the data which is specially suitable for context-aware POI recommendations. In addition, we fuse users' social relations as regularization terms of the factorization to improve the recommendation accuracy. Experimental results on real-world datasets demonstrate the effectiveness of our approach.

As available services accumulate on the Internet, QoS-aware service selection (SSP) becomes an increasingly difficult task. Since Artificial Bee Colony algorithm (ABC) has been successful in solving many problems as a simpler implementation of swarm intelligence, its application to SSP is promising. However, ABC was initially designed for numerical optimization, and its effectiveness highly depends on what we call optimality continuity property of the solution space, i.e., similar variable values (or neighboring solutions) result in similar objective values (or evaluation results). We will show that SSP does not possess such property. We further propose an approximation approach based on greedy search strategies for ABC, to overcome this problem. In this approach, neighboring solutions are generated for a composition greedily based on the neighboring services of its component services. Two algorithms with different neighborhood measures are presented based on this approach. The resulting neighborhood structure of the proposed algorithms is analogical to that of continuous functions, so that the advantages of ABC can be fully leveraged in solving SSP. Also, they are pure online algorithms which are as simple as canonical ABC. The rationale of the proposed approach is discussed and the complexity of the algorithms is analyzed. Experiments conducted against canonical ABC indicate that the proposed algorithms can achieve better optimality within limited time.

Service computing has become a dominant paradigm enabling the building of complex service-oriented systems, with the aim of business added-value. Because these systems are inevitably based on uncontrollable services on the unpredictable Internet, it is important to find effective ways of maximizing the profit of service-oriented systems in such unreliable environment. In this paper, we propose an analytic approach that employs a build-time analysis of the runtime dynamics of service execution to maximize the net profit from delivering composite services under full probability of uncertainty. We also present methods for improving the optimization efficiency, including reusing intermediate computation results and adopting specialized profit optimization algorithms. The superiority of the proposed approach is both theoretically proved and empirically demonstrated through experiments.

Service composition has the capability of constructing coarse-grained solutions by dynamically aggregating a set of services to satisfy complex requirements, but it suffers from dramatic decrease on the efficiency of determining the best composition solution when large scale candidate services are available. Most current approaches look for the optimal composition solution by real-time computation, and the composition efficiency greatly depends on the adopted algorithms. To eliminate such deficiency, this paper proposes a semi-empirical composition approach which incorporates two stages, i.e., periodical clustering and real-time composition. The former partitions the candidate services and historical requirements into clusters based on similarity measurement, and then the probabilistic correspondences between service clusters and requirement clusters are identified by statistical analysis. The latter deals with a new requirement by firstly finding its most similar requirement cluster and the corresponding service clusters by leveraging Bayesian inference, then a set of concrete services are optimally selected from such reduced solution space and constitute the final composition solution. Instead of relying on solely historical data exploration or on pure real-time computation, our approach distinguishes from traditional methods by combining the two perspectives together. Experiments demonstrate the advantages of this approach.

This paper presents a service resource management system. It is oriented to both single and integrated service resources and manages every stage of service resource lifecycle efficiently which includes registering, publication, usage, destruction processes. To satisfy different specific service area, the developer only needs to configure the template provided by the service resource management system, which will greatly shorten the development cycle and increase the development efficiency. The paper is organized as follows: firstly, the paper presents the concept of the service resources, and introduces how to classify and depict them. Secondly, using UML use case analyzes the requirement of the service resource management. Thirdly, the service resource management system design is put forward. At the end, it introduces how to apply the system to a specific service area briefly.

Traditional service composition approaches focus on selecting and composing multiple service components together to fulfill one single requirement. But in most real-world scenarios, there are multiple requirements raised by multiple consumers and they form a discrete and uneven flow (i.e., a temporal sequence). Due to the limited number of available services and their limited capacities, how to ensure the equilibrium between the satisfaction degrees of these temporally sequential requirements becomes an important issue to be addressed. This paper proposes an equilibrium-oriented service composition approach taking into account both the limitedness of service capacity and utilization of historical data. The temporal sequential requirements are divided gradually along with the formation of length-flexible time-segments one by one. Based on this segmentation, service capacity is preserved proportionally for the estimated future requirements, and multiple requirements within one segment are ensured to get relatively equal chances of being satisfied with relatively equal quality. Experiments reveal improved sustainability and superior temporal stability of service quality compared with applying traditional methods to this scenario.