Publisher Copyright: © 2026 Owner/Author.

User-centric recommendation has become essential for delivering personalized services, as it enables systems to adapt to users' evolving behaviors while respecting their long-term preferences and privacy constraints. Although federated learning offers a promising alternative to centralized training, existing approaches largely overlook user behavior dynamics, leading to temporal forgetting and weakened collaborative personalization. In this work, we propose FCUCR, a federated continual recommendation framework designed to support long-term personalization in a privacy-preserving manner. To address temporal forgetting, we introduce a time-aware self-distillation strategy that implicitly retains historical preferences during local model updates. To tackle collaborative personalization under heterogeneous user data, we design an inter-user prototype transfer mechanism that enriches each client's representation using knowledge from similar users while preserving individual decision logic. Extensive experiments on four public benchmarks demonstrate the superior effectiveness of our approach, along with strong compatibility and practical applicability. Code is available at https://github.com/Poizoner/code4FCUCR_www2026.

Federated Recommendation (FR) has emerged as a promising paradigm for addressing the learn-to-rank problem in a privacy-preserving manner. However, effectively incorporating multimodal item features into FR remains an open challenge, due to efficiency constraints, distribution heterogeneity, and feature utilization alignment with the recommendation objective. To tackle these issues, we propose GFMFR, a novel multimodal fusion framework for federated recommendation. Specifically, multimodal representation learning is offloaded to the server, which stores item content and employs a high-capacity encoder to generate expressive representations, thereby alleviating the computational burden on clients. In addition, a group-aware multimodal aggregation mechanism learns shared representations for users with similar interests, enabling knowledge sharing while alleviating distribution heterogeneity. Finally, GFMFR adopts a preference-guided distillation strategy that leverages multimodal information in a way directly aligned with recommendation objectives. The proposed framework can be seamlessly integrated into existing federated recommender systems, enhancing their effectiveness by incorporating multimodal features. Extensive experiments on five benchmark datasets demonstrate that GFMFR consistently outperforms state-of-the-art multimodal FR baselines. The implementation code is available. https://github.com/Zhangwp2420/GFMFR.

Sparse CrowdSensing has emerged as a promising paradigm for data collection, which utilizes mobile devices to gather partial sensing data and infer the remaining ones. Most existing studies focus on data inference. However, they often overlook the impact of unforeseen circumstances. These perturbations can lead to many issues in the data inference process such as over-correction (adjustments are excessively applied), perturbation diffusion (errors affecting related data), and data bias (overall trends in inferred data deviating from the truth). To address the aforementioned challenges, we propose a novel approach called Gradual Matrix Completion (GMC) for handling perturbations in Sparse CrowdSensing. This method distinguishes itself by inferring adjacent unsensed data and using it as a basis for ongoing optimization, gradually completing the entire data matrix. Through the gradual inference process, GMC mitigates perturbation diffusion by continuously refining feature selection, thus preserving and reconstructing essential features. Moreover, GMC concentrates on both intra-area and inter-area spatio-temporal relationships, progressively enhancing its understanding of local and global dependencies. Evaluation of the GMC approach on four diverse real-world datasets highlights its robust capability to complete perturbed data and cope with noise in Sparse CrowdSensing.

Sparse Mobile CrowdSensing (MCS) is a cost-effective paradigm which recruits users to sense only a small amount of data and infer the rest ones. The existing works usually provide rewards for users to cover their sensing costs. Compared with the costs, we actually need to pay for the value of data, which is a new and valuable perspective of data pricing. However, in Sparse MCS, we can hardly estimate the value of data accurately, as it depends not only on the data itself but also on its contribution to inference. Moreover, the sensing data is naturally noisy and its spatiotemporal information may be invisible until we pay for it, which makes the data pricing more challenging. To this end, we investigate the value of sensing data, and propose a data pricing framework for Sparse MCS. We first present the sampling-based uncertainty quantification to evaluate the value of data, with the aim to pay for the intrinsic data values but not the indirect costs. As a byproduct, the noise can be dealt with naturally by such sampling-based methods. Further considering the availability of spatiotemporal information, we also study three cases of data pricing, i.e, paying for sensed data, spatiotemporal information, and next data. Extensive experiments on five real-world data sets verify the effectiveness of our proposals.

Multifaceted user modeling aims to uncover fine-grained patterns and learn representations from user data, revealing their diverse interests and characteristics, such as profile, preference, and personality. Recent studies on foundation model-based recommendation have emphasized the Transformer architecture's remarkable ability to capture complex, non-linear user-item interaction relationships. This paper aims to advance foundation model-based recommendersystems by introducing enhancements to multifaceted user modeling capabilities. We propose a novel Transformer layer designed specifically for recommendation, using the self-attention mechanism to capture sequential user-item interaction patterns. Specifically, we design a group gating network to identify user groups, enabling hierarchical discovery across different layers, thereby capturing the multifaceted nature of user interests through multiple Transformer layers. Furthermore, to broaden the data scope and further enhance multifaceted user modeling, we extend the framework to a federated setting, enabling the use of private datasets while ensuring privacy. Experimental validations on benchmark datasets demonstrate the superior performance of our proposed method.

Spatial transcriptomics (ST) complements scRNA-seq by enabling in situ gene-expression profiling while preserving native tissue architecture and histomorphology. Yet ST data are inherently noisy, high-dimensional, and multimodal, complicating robust clustering. Many existing approaches define neighbor-hoods solely by spatial proximity and aggregate features via naive averaging, which blurs tissue boundaries and underrepre-sents biologically salient heterogeneity and microenvironmental interactions. In this study, we propose MuSACL, a Multimodal Spatially Adaptive Contrastive Learning model that integrates gene expression, spatial coordinates, and histological images to accurately identify spatial domains in ST data. Specifically, MuSACL introduces a Multimodal Proximity Subgraph (MPS) to construct biologically informed graph structures. Furthermore, MuSACL employs heterogeneous modalities using a Graph Attention Network (GAT) and proposes a novel Spatially Adaptive Attention Readout (SAAR) mechanism to enhance contrastive learning with spatial adaptivity. This approach overcomes the limitations of naive feature averaging, yielding more expressive, spatially aware representations. On two public ST datasets, MuSACL significantly surpasses state-of-the-art baselines. Ablations attribute the gains to MPS and SAAR, which enhance representation quality and spatial resolution.

The proliferation of misinformation across diverse social media platforms has drawn significant attention from both academic and industrial communities due to its detrimental effects. Accordingly, automatically distinguishing misinformation, dubbed as Misinformation Detection (MD), has become an increasingly active research topic. The mainstream methods formulate MD as a static learning paradigm, which learns the mapping between the content, links, and propagation of news articles and the corresponding manual veracity labels. However, the static assumption is often violated, since in real-world scenarios, the veracity of news articles may vacillate within the dynamically evolving social environment. To tackle this problem, we propose a novel framework, namely Misinformation detection with Dynamic Environmental Representations (MISDER). The basic idea of MISDER lies in learning a social environmental representation for each period and employing a temporal model to predict the representation for future periods. In this work, we specify the temporal model as the LSTM model, continuous dynamics equation, and pre-trained dynamics system, suggesting three variants of MISDER, namely MISDER-LSTM, MISDER-ODE, and MISDER-PT, respectively. To evaluate the performance of MISDER, we compare it to various MD baselines across 2 prevalent datasets, and the experimental results can indicate the effectiveness of our proposed model.

Graph neural networks have recently demonstrated remarkable performance in predicting material properties. Crystalline material data is manually encoded into graph representations. Existing methods incorporate different attributes into constructing representations to satisfy the constraints arising from symmetries of material structure. However, existing methods for obtaining graph representations are specific to certain constraints, which are ineffective when facing new constraints. In this work, we propose a code generation framework with multiple large language model agents to obtain representations named Rep-CodeGen with three iterative stages simulating an evolutionary algorithm. To the best of our knowledge, Rep-CodeGen is the first framework for automatically generating code to obtain representations that can be used when facing new constraints. Furthermore, a type of representation from generated codes by our framework satisfies six constraints, with codes satisfying three constraints as bases. Extensive experiments on two real-world material datasets show that a property prediction method based on such a graph representation achieves state-of-the-art performance in material property prediction tasks.

Clustered Federated Learning (CFL) addresses the challenges posed by non-IID data by training multiple group- or cluster-specific expert models. However, existing methods often overlook the shared information across clusters, which represents the generalizable knowledge valuable to all participants in the Federated Learning (FL) system. To overcome this limitation, this paper introduces a novel FL framework that distills a universal expert model from the knowledge of multiple clusters. This universal expert captures globally shared information across all clients and is subsequently distributed to each client as the initialization for the next round of model training. The proposed FL framework operates in three iterative steps: (1) local model training at each client, (2) cluster-specific model aggregation, and (3) universal expert distillation. This three-step learning paradigm ensures the preservation of fine-grained non-IID characteristics while effectively incorporating shared knowledge across clusters. Compared to traditional gradient-based aggregation methods, the distillation-based model aggregation introduces greater flexibility in handling model heterogeneity and reduces conflicts among cluster-specific experts. Extensive experimental results demonstrate the superior performance of the proposed method across various scenarios, highlighting its potential to advance the state of CFL by balancing personalized and shared knowledge more effectively.

Property prediction is a fundamental task in crystal material research. To model atoms and structures, structures represented as graphs are widely used and graph learning-based methods have achieved significant progress. Bond angles and bond distances are two key structural information that greatly influence crystal properties. However, most of the existing works only consider bond distances and overlook bond angles. The main challenge lies in the time cost of handling bond angles caused by the large cutoff, which leads to a significant increase in inference time. To address this issue, we first propose a flexible dual-scale neighbor partitioning mechanism with consistent receptive fields in point graphs and edge graphs, which reduces redundant information unrelated to modeling periodicity and improves prediction speed without affecting the modeling of graph periodicity. Then, we propose a novel Atom-Distance-Angle Graph Neural Network (ADA-GNN) for property prediction tasks, which processes node information and structural information separately. The accuracy of predictions and the speed of inference process are both improved with the flexible dual-scale modeling and the specially designed architecture of ADA-GNN. The experimental results validate that our approach achieves state-of-the-art results in two large-scale material benchmark datasets on property prediction tasks.

Sparse Mobile CrowdSensing has emerged as a practical method for data collection, recruiting mobile users to collect partial data and leveraging spatiotemporal correlations to infer the missing data. To improved the QoS of crowdsourced data, existing methods typically assume that the scales of collected data are similar. However, in real-world scenarios, the diversity of user devices results in data collections that vary in scale. More importantly, the collected coarser-scale data points often do not perfectly correspond to multiple finer-scale data points, resulting in highly complex compositional relationships and posing significant challenges for multi-scale data completion. To address these challenges, this paper proposes a multi-scale data completion framework designed to process and integrate multi-scale data with non-aligned compositional relationships. We first align features across scales using the least common multiple scaling, then enhance the interaction and integration of data across scales through a bidirectional processing strategy and modified Mamba architectures, specifically ST-Mamba and Cross-Mamba. Evaluated on six real-world datasets, our study demonstrates the effectiveness of the proposed framework in handling multi-scale data completion challenges, particularly when dealing with non-aligned compositional relationships.

To address the weight coupling problem, certain studies introduced few-shot Neural Architecture Search (NAS) methods, which partition the supernet into multiple sub-supernets. However, these methods often suffer from computational inefficiency and tend to provide suboptimal partitioning schemes. To address this problem more effectively, we analyze the weight coupling problem from a novel perspective, which primarily stems from distinct modules in succeeding layers imposing conflicting gradient directions on the preceding layer modules. Based on this perspective, we propose the Gradient Contribution (GC) method that efficiently computes the cosine similarity of gradient directions among modules by decomposing the Vector-Jacobian Product during supernet backpropagation. Subsequently, the modules with conflicting gradient directions are allocated to distinct sub-supernets while similar ones are grouped together. To assess the advantages of GC and address the limitations of existing Graph Neural Architecture Search methods, which are limited to searching a single type of Graph Neural Networks (Message Passing Neural Networks (MPNNs) or Graph Transformers (GTs)), we propose the Unified Graph Neural Architecture Search (UGAS) framework, which explores optimal combinations of MPNNs and GTs. The experimental results demonstrate that GC achieves state-of-the-art (SOTA) performance in supernet partitioning quality and time efficiency. In addition, the architectures searched by UGAS+GC outperform both the manually designed GNNs and those obtained by existing NAS methods. Finally, ablation studies further demonstrate the effectiveness of all proposed methods.

Low-light image enhancement (LLIE) aims at restoring the color and brightness of underexposed images. Supervised methods suffer from high costs in collecting paired low-normal light images, while unsupervised approaches require intricate loss functions. To tackle these dual challenges, we propose the Trouble-Maker Learning (TML) strategy, which leverages images with normal light as training inputs. TML comprises two core components. Firstly, the Troublemaker Model (TM) generates pseudo low-light images from normal images, thereby alleviating the need for pairwise data and reducing associated costs. Secondly, the Predicting Model (PM) enhances the brightness of pseudo low-light images. Additionally, we integrate an Enhancing Model (EM) to further refine the visual quality of the PM's outputs. In LLIE tasks, it is crucial to capture global element correlations, as this allows for the extraction of more information pertaining to the same object. Convolutional Neural Networks (CNNs) and self-attention mechanisms are not well-suited to this task due to the local CNN operators, and high time complexity, respectively. To address these limitations, we propose Global Dynamic Convolution (GDC) with a time complexity of O(n). Essentially, GDC mimics the partial calculation process of self-attention to establish element-wise correlations. Building upon the GDC module, we develop the UGDC model. Finally, we explore the application of Data Fusion in the field of LLIE. Based on the Retinex theory, we conducted feature-level fusion using low-light images, illumination components and reflection components, which further enhance the performance of the LLIE system. Extensive quantitative and qualitative experiments demonstrate that UGDC, trained with TML and via data fusion, can achieve performance competitive with state-of-the-art approaches on public datasets. The source code of this paper is publicly available at https://github.com/Rainbowman0/TML_LLIE, facilitating reproducibility of the research findings.

Temporal non-stationarity, the phenomenon that time series distributions change over time, poses fundamental challenges to reliable time series forecasting. Intuitively, the complex time series can be decomposed into two factors, i.e., time-invariant and time-varying components, which indicate static and dynamic patterns, respectively. Nonetheless, existing methods often conflate the time-varying and time-invariant components, and jointly learn the combined long-term patterns and short-term fluctuations, leading to suboptimal performance facing distribution shifts. To address this issue, we initiatively propose a lightweight static-dynamic decomposition framework, TimeEmb, for time series forecasting. TimeEmb innovatively separates time series into two complementary components: (1) time-invariant component, captured by a novel global embedding module that learns persistent representations across time series, and (2) time-varying component, processed by an efficient frequency-domain filtering mechanism inspired by full-spectrum analysis in signal processing. Experiments on real-world datasets demonstrate that TimeEmb outperforms state-of-the-art baselines and requires fewer computational resources. We conduct comprehensive quantitative and qualitative analyses to verify the efficacy of static-dynamic disentanglement. This lightweight framework can also improve existing time-series forecasting methods with simple integration. To ease reproducibility, our code is available at https://github.com/showmeon/TimeEmb.

Knowledge graphs (KGs) consist of well-organized external information and have been proven to enhance recommendation quality effectively. Most KG-aware recommender systems are developed using real number space embeddings. In recent years, learning representations in the hypercomplex space has gained success and attention. Compared to single-component real-valued vectors, multi-component hypercomplex embeddings offer greater expressiveness, facilitating more meaningful modeling of users, items, entities, and their relations in the user-item interaction graph and KG. In this paper, we explore the integration of hypercomplex algebras in KG-aware recommendation and propose a Hypercomplex Knowledge Graph-aware Recommender (HKGR) method. Our HKGR models the interaction graph and KG in the hypercomplex space by utilizing specially designed hypercomplex graph neural networks. In particular, HKGR employs a hypercomplex attention-based aggregator to capture the structure and semantics of the KG. In the recommendation prediction phase, we design a hypercomplex interaction network that can approximate the high-order component interactions between users and items. Furthermore, we introduce a hypercomplex contrastive learning operator to strengthen cooperative signals between the interaction graph and KG modelings. Experiment results on the four real-world datasets show that our HKGR outperforms the state-of-the-art recommender baselines.

Graph-based social recommender systems utilize user-item interaction graphs and user-user social graphs to model user preferences. However, their performance can be limited by redundant and noisy information in these two graphs. Although several recommender studies on data denoising exist, most either rely on heuristic assumptions, which limit their adaptability, or use a single model that combines denoising and recommendation, potentially imposing substantial demands on the model capacity. To address these issues, we propose a dual Graph Denoising Social Recommender (GDSR), which consists of two steps: graph denoising and user preference prediction. First, we design a denoising module which exploits a dual diffusion model to alleviate noises in the interaction and social graphs by performing multi-step noise diffusion and removal. We develop three kinds of conditions to guide our dual graph diffusion paradigm and propose a cross-domain signal guidance mechanism to enhance the structure of denoised graphs. Second, we devise a recommender module that employs a dual graph learning structure on denoised graphs to generate recommendations. Moreover, we use additional supervision signals from the diffusion-enhanced data augmentation to introduce a graph contrastive learning task, enhancing the recommender module's representation quality and robustness. Experiment results show the effectiveness of our GDSR.

Out-of-distribution (OOD) generalization is critical for applying reinforcement learning algorithms to real-world applications. To address the OOD problem, recent works focus on learning an OOD adaptation policy by capturing the causal factors affecting the environmental dynamics. However, these works recover the causal factors with only an entangled or binary form, resulting in a limited generalization of the policy that requires extra data from the testing environments. To break this limitation, we propose generalizable causal reinforcement learning (GCRL) to learn a disentangled representation of causal factors, on the basis of which we learn a policy that achieves the OOD generalization without extra training. For capturing the causal factors, GCRL deploys a weakly supervised signal with a two-stage constraint to ensure that all factors can be disentangled. Then, to achieve the OOD generalization through causal factors, we establish the dependence of actions on the learned representation and optimize the policy model across multiple environments. Experimental results show that the established dependence recovers the correct relationship between causal factors and actions when the learned policy could address the target tasks in training environments. Benefiting from the recovered relationship, GCRL achieves the OOD generalization on eight benchmarks from Causal World and Mujoco. Moreover, the policy learned by our model is more explainable and can be controlled to generate semantic actions by intervening in the representation of causal factors.

Complex networks enable to represent and characterize the interactions between entities in various complex systems which widely exist in the real world and usually generate vast amounts of data about all the elements, their behaviors and interactions over time. The studies concentrating on new network analysis approaches and methodologies are vital because of the diversity and ubiquity of complex networks. The stochastic block model (SBM), based on Bayesian theory, is a statistical network model. SBMs are essential tools for analyzing complex networks since SBMs have the advantages of interpretability, expressiveness, flexibility and generalization. Thus, designing diverse SBMs and their learning algorithms for various networks has become an intensively researched topic in network analysis and data mining. In this article, we review, in a comprehensive and in-depth manner, SBMs for different types of networks (i.e., model extensions), existing methods (including parameter estimation and model selection) for learning optimal SBMs for given networks and SBMs combined with deep learning. Finally, we provide an outlook on the future research directions of SBMs.

Federated recommender systems (FedRecSys) have emerged as a pivotal solution for privacy-aware recommendations, balancing growing demands for data security and personalized experiences. Current research efforts predominantly concentrate on adapting traditional recommendation architectures to federated environments, optimizing communication efficiency, and mitigating security vulnerabilities. However, user personalization modeling, which is essential for capturing heterogeneous preferences in this decentralized and non-IID data setting, remains underexplored. This survey addresses this gap by systematically exploring personalization in FedRecSys, charting its evolution from centralized paradigms to federated-specific innovations. We establish a foundational definition of personalization in a federated setting, emphasizing personalized models as a critical solution for capturing fine-grained user preferences. The work critically examines the technical hurdles of building personalized FedRecSys and synthesizes promising methodologies to meet these challenges. As the first consolidated study in this domain, this survey serves as both a technical reference and a catalyst for advancing personalized FedRecSys research.

Mobile CrowdSensing (MCS) has emerged as a popular paradigm to engage mobile users in collaborative sensing tasks. However, its performance is hindered by its limited spatiotemporal range and the cost of data collection. An effective strategy is to integrate Sparse MCS with data completion, allowing for unsensed data inference. However, when confronted with situations where sensed data is excessively sparse, data inference results may be unsatisfactory due to several challenges including: 1) uneven data distribution, 2) complex spatiotemporal correlation, and 3) the presence of inference noise. To address these challenges, we propose a model named Spatiotemporal Evolutionary Inference (STEI) that achieves accurate inference of unsensed data in Sparse MCS. Specifically, we complete the unsensed data by uncovering strong local correlations in the data and gradually evolving those correlations to the global situation. In each evolution step, we thoroughly consider the impact of spatiotemporal consistency and difference. To minimize the interference of noise during the evolution process, we design an adaptive coefficient to enhance the dependence on sensed data. Finally, to validate the effectiveness of STEI, we conduct extensive qualitative and quantitative experiments using three popular datasets. The experimental results demonstrate that our approach excels in accurately inferring data, particularly in situations where the distribution of data is notably uneven.

Mobilecrowdsensing (MCS) is a powerful technique that enables a variety of urban tasks, including temperature monitoring, location-based services, and urban path recommendations. However, these tasks often face the challenge of sparse and incomplete sensing data, undermining their effectiveness and reliability.Sparsedatacompletion (SDC) methods have been developed to infer missing or unobserved data by leveraging spatio-temporal correlations to tackle this issue. This forms the core concept of thesparsemobilecrowdsensing problem (SMCS), which aims to improve the performance of downstream tasks through inferred data. Despite the potential benefits, most existing SMCS methods fail to consider the trade-off between the cost of SDC and the benefits for downstream tasks. These methods often treat SDC and downstream tasks as independent modules, resulting in suboptimal outcomes. In this paper, we investigate the impact of SDC on the SMCS paradigm, both qualitatively and quantitatively. We establish the upper bound of performance achievable when applying SDC in SMCS under different levels of sensing data sparsity. Based on these studies and findings, we propose a practical and flexible framework calledSDC-EVA,SensingDataCompletionEVAluation framework. This framework allows for applying different SDC methods in SMCS, considering factors such as computing complexity, storage space, and associated costs. Our proposed framework allows researchers to assess the necessity and feasibility of integrating SDC into SMCS systems before designing and deploying them in real-world scenarios. This assessment can be tailored to specific data sparsity and contextual information. To validate the effectiveness of our proposed evaluation framework, we conduct experiments in various real-world scenarios involving different combinations of SDC and downstream tasks. The results demonstrate the superiority of our framework in improving the performance of SMCS. By presenting these findings, we aim to contribute to developing SMCS techniques and provide valuable insights for researchers and practitioners.

Multiagent policy gradients (MAPGs), an essential branch of reinforcement learning (RL), have made great progress in both industry and academia. However, existing models do not pay attention to the inadequate training of individual policies, thus limiting the overall performance. We verify the existence of imbalanced training in multiagent tasks and formally define it as an imbalance between policies (IBPs). To address the IBP issue, we propose a dynamic policy balance (DPB) model to balance the learning of each policy by dynamically reweighting the training samples. In addition, current methods for better performance strengthen the exploration of all policies, which leads to disregarding the training differences in the team and reducing learning efficiency. To overcome this drawback, we derive a technique named weighted entropy regularization (WER), a team-level exploration with additional incentives for individuals who exceed the team. DPB and WER are evaluated in homogeneous and heterogeneous tasks, effectively alleviating the imbalanced training problem and improving exploration efficiency. Furthermore, the experimental results show that our models can outperform the state-of-the-art MAPG methods and boast over 12.1% performance gain on average.

Recently, there has been a surge in recommendations based on heterogeneous information networks (HINs), attributed to their ability to integrate complex and rich semantics. Despite this advancement, most HIN-based recommenders overlook two critical aspects. First, they often fail to consider HIN's heterophily nature, hindering the capture of non-local structures in HINs. Second, most methods lack the capability for logical reasoning. In this paper, we propose a novel structure- and logic-aware heterogeneous graph learning framework for recommender systems (SLHRec). Our SLHRec contains a structure-aware module and a logic-aware module. The former uses network geometry to construct non-local neighborhoods for nodes in HINs, and then introduces a graph neural network to integrate constructed neighbors for modeling the heterophily of HINs. The logic-aware module uses the Markov logic network (MLN) to infuse logic rules into heterogeneous graph learning, thereby boosting logic reasoning in recommendations. Furthermore, we utilize contrastive learning to model cooperative signals between modules, enabling them to complement each other. In the prediction stage, both modules contribute to generating recommendations. Compared with several strong recommender baselines, our SLHRec achieves superior performance on four real-world datasets.

In-context learning is a promising approach for offline reinforcement learning (RL) to handle online tasks, which can be achieved by providing task prompts. Recent works demonstrated that in-context RL could emerge with self-improvement in a trial-and-error manner when treating RL tasks as an across-episodic sequential prediction problem. Despite the self-improvement not requiring gradient updates, current works still suffer from high computational costs when the across-episodic sequence increases with task horizons. To this end, we propose an In-context Decision Transformer (IDT) to achieve self-improvement in a high-level trial-and-error manner. Specifically, IDT is inspired by the efficient hierarchical structure of human decision-making and thus reconstructs the sequence to consist of high-level decisions instead of low-level actions that interact with environments. As one high-level decision can guide multi-step low-level actions, IDT naturally avoids excessively long sequences and solves online tasks more efficiently. Experimental results show that IDT achieves state-of-the-art in long-horizon tasks over current in-context RL methods. In particular, the online evaluation time of our IDT is 36$\times$ times faster than baselines in the D4RL benchmark and 27$\times$ times faster in the Grid World benchmark.

Mobile Crowdsensing is a type of technology that utilizes mobile devices and volunteers to gather data about specific topics at large scales in real-time. However, in practice, limited participation leads to missing data, i.e., the collected data may be sparse, which makes it difficult to perform accurate analysis. A possible technique called sparse crowdsensing incorporates the sparse case with data completion, where unsensed data could be estimated through inference. However, sparse crowdsensing typically suffers from poor performance during the data completion stage due to various challenges: the sparsity of the sensed data, reliance on numerous timeslots, and uncertain spatiotemporal connections. To resolve such few-shot issues, the proposed solution uses the Correlated Data Fusion for Matrix Completion (CDFMC) approach, which leverages a small amount of objective data to retrain an auxiliary dataset-based pre-trained model that can estimate unsensed data efficiently. CDFMC is trained using a combination of the traditional Deep Matrix Factorization and the Kalman Filtering, which not only enables the efficient representation and comparison of data samples but also fuses the objective data and auxiliary data effectively. Evaluation results show that the proposed CDFMC outperforms baseline techniques, achieving high accuracy in completing unsensed data with minimal training data.

The federated recommendation system is an emerging AI service architecture that provides recommendation services in a privacy-preserving manner. Using user-relation graphs to enhance federated recommendations is a promising topic. However, it is still an open challenge to construct the user-relation graph while preserving data locality-based privacy protection in federated settings. Inspired by a simple motivation, similar users share a similar vision (embeddings) to the same item set, this paper proposes a novel Graph-guided Personalization for Federated Recommendation (GPFedRec). The proposed method constructs a user-relation graph from user-specific personalized item embeddings at the server without accessing the users' interaction records. The personalized item embedding is locally fine-tuned on each device, and then a user-relation graph will be constructed by measuring the similarity among client-specific item embeddings. Without accessing users' historical interactions, we embody the data locality-based privacy protection of vanilla federated learning. Furthermore, a graph-guided aggregation mechanism is designed to leverage the user-relation graph and federated optimization framework simultaneously. Extensive experiments on five benchmark datasets demonstrate GPFedRec's superior performance. The in-depth study validates that GPFedRec can generally improve existing federated recommendation methods as a plugin while keeping user privacy safe. Code is available https://github.com/Zhangcx19/GPFedRec

Recent works have shown the remarkable superiority of transformer models in reinforcement learning (RL), where the decision-making problem is formulated as sequential generation. Transformer-based agents could emerge with self-improvement in online environments by providing task contexts, such as multiple trajectories, called in-context RL. However, due to the quadratic computation complexity of attention in transformers, current in-context RL methods suffer from huge computational costs as the task horizon increases. In contrast, the Mamba model is renowned for its efficient ability to process long-term dependencies, which provides an opportunity for in-context RL to solve tasks that require long-term memory. To this end, we first implement Decision Mamba (DM) by replacing the backbone of Decision Transformer (DT). Then, we propose a Decision Mamba-Hybrid (DM-H) with the merits of transformers and Mamba in high-quality prediction and long-term memory. Specifically, DM-H first generates high-value sub-goals from long-term memory through the Mamba model. Then, we use sub-goals to prompt the transformer, establishing high-quality predictions. Experimental results demonstrate that DM-H achieves state-of-the-art in long and short-term tasks, such as D4RL, Grid World, and Tmaze benchmarks. Regarding efficiency, the online testing of DM-H in the long-term task is 28$\times$ times faster than the transformer-based baselines.

Zero-shot image classification, which aims to predict unseen classes whose samples have never appeared during the training phase, is crucial in the Web domain because many new web images appear on various websites. Attributes, as annotations for class-level characteristics, are widely used semantic information for this task. However, most current methods often fail to capture discriminative image features between similar images from different classes, leading to unsatisfactory zero-shot image classification results. This is because they solely focus on limited visual-attribute feature alignment. Therefore, we propose a Zero-Shot image Classification with Logic adapter and Rule prompt method called ZSCLR, which utilizes logic adapter and rule prompts to encourage the model to capture discriminative image features and achieve reasoning. Specifically, ZSCLR consists of a visual perception module and a logic adapter. The visual perception module extracts image features from training data. At the same time, the logic adapter utilizes the Markov logic network to encode the extracted image features and rule prompts for refining the discriminative image features. Due to predicates of rule prompts representing symbolic discriminative features, the proposed model can focus more on these discriminative features and achieve more precise image classification. Additionally, the logic adapter enables the model to adapt from recognizing images in seen classes to those in unseen classes through the reasoning of the Markov logic networks. We implement experiments on three standard zero-shot image classification benchmarks, and ZSCLR achieves competitive performance. Furthermore, ZSCLR can provide explanations for its predictions through rule prompts.

Federated recommendation system usually trains a global model on the server without direct access to users' private data on their own devices. However, this separation of the recommendation model and users' private data poses a challenge in providing quality service, particularly when it comes to new items, namely cold-start recommendations in federated settings. This paper introduces a novel method called Item-aligned Federated Aggregation (IFedRec) to address this challenge. It is the first research work in federated recommendation to specifically study the cold-start scenario. The proposed method learns two sets of item representations by leveraging item attributes and interaction records simultaneously. Additionally, an item representation alignment mechanism is designed to align two item representations and learn the meta attribute network at the server within a federated learning framework. Experiments on four benchmark datasets demonstrate IFedRec's superior performance for cold-start scenarios. Furthermore, we also verify IFedRec owns good robustness when the system faces limited client participation and noise injection, which brings promising practical application potential in privacy-protection enhanced federated recommendation systems. The implementation code is available

We study defending strategies against adversarial attacks onCombinatorial Multi-Armed Bandits(CMAB) algorithms. CMAB is an effective sequence decision making tool that has been broadly applied in online real-world applications. We consider a realistic CMAB setting, budgeted CMAB, in which multiple arms associated with pulling costs and unknown rewards are pulled per round, aiming to maximize the cumulative reward under a budget constraint. However, the adversarial attack against budgeted CMAB is rarely studied, posing a very important security issue. Specifically, a suboptimal arm that is not pulled (i.e., attacker) can hijack the budgeted CMAB algorithm's behavior, forcing itself to be pulled frequently by manipulating other arms' rewards. Existing strategies cannot prevent such attacks. Motivated by this, we closely study the adversarial attack against a popular budgeted CMAB algorithm, exposing a significant security threat to real-world applications. The attack extends to other algorithms with certain customization. To address this, we incorporate a truthful pricing-based defending strategy that prevents such attacks effectively and ensures arms share pulling costs truthfully. Extensive simulations have illustrated the proposed attack strategy can hijack the algorithm efficiently, while the defending strategy provides attack prevention, individual rationality, and asymptotic truthfulness guarantees.

Parking violation is a common urban problem in major cities all over the world. Traditional approaches for detecting parking violations mainly rely on fixed deployed sensors and enforcement agencies, which suffer from high deployment costs and limited coverage. With the rapid development of mobile networks, Mobile CrowdSensing (MCS) has been an effective sensing paradigm. The crowdsensing data can help predict the future parking violation distribution, and the prediction results can provide guidance for user scheduling, i.e., sending the mobile users to patrol areas where many parking violation events may occur. Inspired by this idea, we propose a comprehensive data-driven crowdsensing framework, which incorporates the nested design of a generative model for spatial-temporal data and a user scheduling model. The generative model extracts parking violation hotspots via a data completion module and violation prediction module. Since crowdsensing data is usually temporally sparse and unevenly distributed, a data completion module is proposed to infer the missing statistics in unsensed areas. The violation prediction module then predicts the parking violation distribution. Given the predicted results, the deep reinforcement learning-based user scheduling model coordinates users to visit hotspots for violation detection. Iteratively, the newly collected data can be used to predict the future violation distribution. Finally, we conduct extensive simulations based on two real-world datasets from two large urban cities. The simulation verifies the prediction accuracy and scheduling effectiveness of the proposed framework compared with the baselines.

Mobile CrowdSensing (MCS) is a widely adopted approach that involves engaging mobile users to collaboratively perform diverse sensing tasks. In Sparse CrowdSensing, the completion of data from partially-sensed sources plays a pivotal role in urban sensing applications. This process is essential as it enables efficient data representation, enhances urban analysis capabilities, and ultimately facilitates informed city planning decisions. By leveraging the power of mobile users, Sparse CrowdSensing contributes to the comprehensive understanding of urban environments, enabling effective utilization of data for optimizing urban infrastructure and fostering sustainable urban development. To achieve accurate completion results, previous methods usually utilize the universal similarity and conventional tendency while incorporating only a single dataset to infer the full map. However, in real-world scenarios, there may exist many kinds of data (inter-data), that could help to complement each other. Moreover, for each kind of data (intra-data), there usually exists a few but important spatiotemporal fracture data which caused by the special events (e.g. data loss, equipment failure, etc.), which may behave in a different way as the statistical patterns. Thus, it is an essential task to consider spatiotemporal fracture data caused by the special cases in spatiotemporal data inference, especially using both intra- and inter-data, because of the following challenges: 1) the sparsity of the sensed data, 2) the complex spatiotemporal relations and 3) the uncertain scale of a spatiotemporal fracture. To this end, focusing on the large-scale spatiotemporal fracture, we propose a data completion method that exploits both intra- and inter-data correlations for enhancing performance. Specifically, for the purpose of generating spatiotemporal fracture data, there isStackedGenerativeMatrixCompletion(SGMC)by combining previous Stacked Deep Matrix Factorization (SDMF) and Generative Adversarial Networks (GANs), which improves a lot. Along this line, we extract the features of spatiotemporal data and further efficiently complete and predict the unsensed data. Finally, we conduct both qualitative and quantitative experiments on two different datasets, and the results demonstrate that the performance of our method outperforms the state-of-the-art baselines.

Fault Scenario Identification (FSI) is a challenging task that aims to automatically identify the fault types in communication networks from massive alarms to guarantee effective fault recoveries. Existing methods are developed based on rules, which are not accurate enough due to the mismatching issue. In this paper, we propose an effective method named Knowledge-Enhanced Graph Neural Network (KE-GNN), the main idea of which is to integrate the advantages of both the rules and GNN. This work is the first work that employs GNN and rules to tackle the FSI task. Specifically, we encode knowledge using propositional logic and map them into a knowledge space. Then, we elaborately design a teacher-student scheme to minimize the distance between the knowledge embedding and the prediction of GNN, integrating knowledge and enhancing the GNN. To validate the performance of the proposed method, we collected and labeled three real-world 5G fault scenario datasets. Extensive evaluation conducted on these datasets indicates that our method achieves the best performance compared with other representative methods, improving the accuracy by up to 8.10%. Furthermore, the proposed method achieves the best performance against a small dataset setting and can be effectively applied to a new carrier site with a different topology structure.

As deep learning gains popularity in modelling dynamical systems, we expose an underappreciated misunderstanding relevant to modelling dynamics on networks. Strongly influenced by graph neural networks, latent vertex embeddings are naturally adopted in many neural dynamical network models. However, we show that embeddings tend to induce a model that fits observations well but simultaneously has incorrect dynamical behaviours. Recognising that previous studies narrowly focus on short-term predictions during the transient phase of a flow, we propose three tests for correct long-term behaviour, and illustrate how an embedding-based dynamical model fails these tests, and analyse the causes, particularly through the lens of topological conjugacy. In doing so, we show that the difficulties can be avoided by not using embedding. We propose a simple embedding-free alternative based on parametrising two additive vector-field components. Through extensive experiments, we verify that the proposed model can reliably recover a broad class of dynamics on different network topologies from time series data.

Federated recommendation is a new Internet service architecture that aims to provide privacy-preserving recommendation services in federated settings. Existing solutions are used to combine distributed recommendation algorithms and privacy-preserving mechanisms. Thus it inherently takes the form of heavyweight models at the server and hinders the deployment of on-device intelligent models to end-users. This paper proposes a novel Personalized Federated Recommendation (PFedRec) framework to learn many user-specific lightweight models to be deployed on smart devices rather than a heavyweight model on a server. Moreover, we propose a new dual personalization mechanism to effectively learn fine-grained personalization on both users and items. The overall learning process is formulated into a unified federated optimization framework. Specifically, unlike previous methods that share exactly the same item embeddings across users in a federated system, dual personalization allows mild finetuning of item embeddings for each user to generate user-specific views for item representations which can be integrated into existing federated recommendation methods to gain improvements immediately. Experiments on multiple benchmark datasets have demonstrated the effectiveness of PFedRec and the dual personalization mechanism. Moreover, we provide visualizations and in-depth analysis of the personalization techniques in item embedding, which shed novel insights on the design of recommender systems in federated settings. The code is available.

Actor-critic framework has achieved tremendous success in a great many of decision-making scenarios. Nevertheless, when updating the value of new states and actions in the long-term scene, these methods suffer from misestimate problem and gradient variance problem, significantly reducing convergence speed and robustness of the policy. These problems severely limit the application scope of these methods. In this paper, we first proposed QVDDPG, a deep RL algorithm based on the iterative target value update process. The QV learning method alleviates the problem of misestimate by making use of the guidance of Q value and the fast convergence of V value, thus accelerating the convergence speed. In addition, the actor utilizes a constrained balanced gradient and establishes a hidden state for the continuous action space network for the sake of robustness of the model. We give the update relation among the value functions and the constraint conditions of gradient estimation. We measure our method on the PyBullet and achieved state-of-the-art performance. Moreover, we demonstrate that, our method has higher robustness and convergence speed across different tasks compared to other algorithms.

In visual-based Reinforcement Learning (RL), agents often struggle to generalize well to environmental variations in the state space that were not observed during training. The variations can arise in both task-irrelevant features, such as background noise, and task-relevant features, such as robot configurations, that are related to the optimal decisions. To achieve generalization in both situations, agents are required to accurately understand the impact of changed features on the decisions, i.e., establishing the true associations between changed features and decisions in the policy model. However, due to the inherent correlations among features in the state space, the associations between features and decisions become entangled, making it difficult for the policy to distinguish them. To this end, we propose Saliency-Guided Features Decorrelation (SGFD) to eliminate these correlations through sample reweighting. Concretely, SGFD consists of two core techniques: Random Fourier Functions (RFF) and the saliency map. RFF is utilized to estimate the complex non-linear correlations in high-dimensional images, while the saliency map is designed to identify the changed features. Under the guidance of the saliency map, SGFD employs sample reweighting to minimize the estimated correlations related to changed features, thereby achieving decorrelation in visual RL tasks. Our experimental results demonstrate that SGFD can generalize well on a wide range of test environments and significantly outperforms state-of-the-art methods in handling both task-irrelevant variations and task-relevant variations.

Variational graph autoencoder (VGAE) is a promising deep probabilistic model in graph representation learning. However, most existing VGAEs adopt the mean-field assumption, and cannot characterize the graphs with noise well. In this paper, we propose a novel deep probabilistic model for graph analysis, termed Multi-head Variational Graph Autoencoder Constrained by Sum-product Networks (named SPN-MVGAE), which helps to relax the mean-field assumption and learns better latent representation with fault tolerance. Our proposed model SPN-MVGAE uses conditional sum-product networks as constraints to learn the dependencies between latent factors in an end-to-end manner. Furthermore, we introduce the superposition of the latent representations learned by multiple variational networks to represent the final latent representations of nodes. Our model is the first use sum-product networks for graph representation learning, extending the scope of sum-product networks applications. Experimental results show that compared with other baseline methods, our model has competitive advantages in link prediction, fault tolerance, node classification, and graph visualization on real datasets.

Recently, deep learning techniques have yielded immense success on recommender systems. However, one weakness of most deep methods is that, users/items mutual semantic relationships, which are latent in the user-item interactions, are not distilled out explicitly. Moreover, most methods have been primarily focused on representation learning in euclidean geometry. Since recent studies have shown that the bipartite graph structure has the non-euclidean latent anatomy, euclidean embeddings may suffer from a certain degree of distortion. In this work, we presentHyperbolicNeuralCollaborativeRecommender (HNCR), a deep hyperbolic representation learning method that exploits mutual semantic relationships among users/items for collaborative filtering tasks. HNCR first introduces a neighbor construction strategy to build user and item semantic neighborhoods. Then HNCR develops a framework based on deep learning and hyperbolic geometry to integrate constructed neighborhoods into recommendation. To evaluate our method, we conduct experiments on the four datasets. Experimental results show the superiority of HNCR compared with its euclidean counterpart and state-of-the-art recommendation baselines. The results also indicate that hyperbolic representations can reflect meaningful data insights.

Mobile CrowdSensing (MCS) has recently become a promising data acquisition paradigm, which recruits a large number of users to collect data from the target sensing areas. Obviously, with the increase of sensing scale and the decrease of sensing granularity, traditional MCS cannot fully cover the required sensing areas especially the inaccessible areas. As a variant, Sparse MCS can utilize the spatiotemporal correlations in sensing data to infer the whole sensing map only by sensing a few subareas. However, in many real-world scenarios, such as traffic congestion prediction or parking occupancy detection, inferring the current unsensed data may not be the final goal. By comparison, it is more important to get the future information through the sparse sensed data. In this paper, we turn attention from inferring the current unsensed data to predicting the future unknown data and propose an urban inference and prediction framework in Sparse MCS. To deal with the sparse sensed data, we first present a bipartite-graph-based matrix completion algorithm with spatiotemporal constraints to accurately recover the current full map. Then, by exploiting spatiotemporal correlations based on the inferred full map, we present a Graph Convolutional Networks (GCN) with spatiotemporal attention to predict the future maps. Furthermore, we design a spatiotemporal iterative method to repeatedly update the spatiotemporal attentions and constraints, in order to connect the urban inference and prediction to improve the accuracy of the whole framework. Extensive experiments have been conducted on two types of typical urban sensing tasks with four real-world data sets, which verify the effectiveness of our proposed algorithms in improving the inference and prediction accuracy with the sparse sensed data.

Mobile CrowdSensing (MCS) is a popular data collection paradigm which usually faces the problem of sparse sensed data because of the limited sensing cost. In order to address the situation of sparse data, sparse MCS recruits users to sense important areas and infers completed data by data completion, which is crucial in sparse MCS for urban sensing applications (e.g. enhancing data expression, improving urban analysis, guiding city planning, etc.) To achieve accurate completion results, previous methods usually utilize the universal similarity and conventional tendency while incorporating only a single dataset to infer the full map. However, in real-world scenarios, there may exist many kinds of data (inter-data), that could help to complement each other. Moreover, for each kind of data (intra-data), there usually exist a few but important outliers caused by the special events (e.g., parking peak, traffic congestion, or festival parade), which may behave in a different way as the statistical patterns. These outliers cannot be ignored, while it is difficult to detect and recover them in data completion because of the following challenges: 1) the infrequency and unpredictability of outliers’ occurrence, 2) the large deviations against the means compared to normal values, and 3) the complex spatiotemporal relations among inter-data. To this end, focusing on spatiotemporal data with both intra- and inter-data correlations, we propose a matrix completion method that takes outliers’ effects into consideration and exploits both intra- and inter-data correlations for enhancing performance. Specifically, we first conduct the Deep Matrix Factorization (DMF) with multiple auxiliary Neural Networks, which named Stacked Deep Matrix Factorization (SDMF). Note that the loss function of SDMF is no longer the previous MSE loss function, but replaced with an Outlier Value Loss (OVL) function to effectively detect and recover the outliers. Moreover, a spatiotemporal outlier value memory network is added for further enhancing the outlier inference. Finally, we take extensive qualitative and quantitative experiments on two popular datasets each with two types of sensing data, and the experimental results indicate the advantages of our method that outperforms the state-of-the-art methods.

This paper aims to leverage symbolic knowledge to improve the performance and interpretability of the Visual Relationship Detection (VRD) models. Existing VRD methods based on deep learning suffer from the problems of poor performance on insufficient labeled examples and lack of interpretability. To overcome the aforementioned weaknesses, we integrate symbolic knowledge into deep learning models and propose a bi-level probabilistic graphical reasoning framework called BPGR. Specifically, in the high-level structure, we take the objects and relationships detected by the VRD model as hidden variables (reasoning results); In the low-level structure of BPGR, we use Markov Logic Networks (MLNs) to project First-Order Logic (FOL) as observed variables (symbolic knowledge) to correct error reasoning results. We adopt a variational EM algorithm for optimization. Experiments results show that our BPGR improves the performance of the VRD models. In particular, BPGR can also provide easy-to-understand insights for reasoning results to show interpretability.

Fault localization, as a crucial process in network fault management, is the process of deducing the exact source of a failure from a sequence of observed symptoms. Existing methods for this task are either expert system-based or data-driven. However, as communication networks grow and become more complex, conventional expert system-based approaches face problems of inefficiency and inflexibility. Besides, purely data-driven machine learning algorithms are not widely accepted in the industry because of their demand for large training sets and lack of explainability. Inspired by the dual process theory in psychology, we propose a dual-system method, named DualSys, for fault localization in this paper. In the proposed method, a fast data-driven intuitive system and a slow knowledge-driven logical system cooperate sequentially to fulfill the task. To avoid possible conflicts between the two systems, we further propose two conflict-easing mechanisms and incorporate them into the overall process. We validate our method using data from a real-world communication network. Experiment results indicate that our proposed method can get the same accuracy and explainability as knowledge-based approaches and achieve higher efficiency. As a result, we argue that our method provides network operators with a promising choice for efficient fault localization.

Hypercomplex algebras are well-developed in the area of mathematics. Recently, several hypercomplex recommendation approaches have been proposed and yielded great success. However, two vital issues have not been well-considered in existing hypercomplex recommenders. First, these methods are only designed for specific and low-dimensional hypercomplex algebras (e.g., complex and quaternion algebras), ignoring the exploration and utilization of high-dimensional ones. Second, most recommenders treat every user-item interaction as an isolated data instance, without considering high-order collaborative relationships.

With the development of communications, networking, and information technology, Crowdsensed Data Trading (CDT) becomes a novel data trading paradigm. In CDT, the data requesters publish crowdsensing tasks with specific data requirements, and then workers complete these tasks, upload the data and obtain corresponding rewards. To efficiently deal with data trading, most of the existing CDT systems assume a trusted centralized platform. However, we argue that the platform may collude with workers or requesters to trick others for achieving more benefits. For example, according to the workers’ uploaded data, the platform can modify the reward functions by colluding with the requester. Similarly, the platform might collude with workers to let them know the reward function, then workers could forge data. Meanwhile, requesters and workers may also be malicious. For example, requesters may post tasks but fail to pay and workers can upload wrong data to mislead the system. To solve the above problems, we combine the Crowdsensed Data Trading system with intelligent Blockchain (CDT-B), which contains a smart contract called CDToken. As a credible third-party, the CDToken is used to record the requesters’ reward function and workers’ data uploading function to avoid targeted trick. At the same time, we not only design a Data Uploading and Preprocessing (DUP) mechanism in CDToken to collect and process the workers’ sensed data, but also propose a Grouping Truth Discovery (GTD) to evaluate their data quality for determining the payments. Moreover, to hold a large number of requesters and workers in CDT-B, we propose a Layered Sharding blockchain based on Membership Degree (LSMD) to solve the blockchain inefficiency problem. Finally, we deploy CDToken to an experimental environment based on Ethereum and demonstrate its efficient performance and practicability.

The key to the effective control of a diffusion system lies in how accurately we could predict its unfolding dynamics based on the observation of its current state. However, in the real-world applications, it is often infeasible to conduct a timely and yet comprehensive observation due to resource constraints. In view of such a practical challenge, the goal of this work is to develop a novel computational method for performing active observations, termed active surveillance, with limited resources. Specifically, we aim to predict the dynamics of a large spatio-temporal diffusion system based on the observations of some of its components. Towards this end, we introduce a novel measure, the γ value, that enables us to identify the key components by means of modeling a sentinel network with a row sparsity structure. Having obtained a theoretical understanding of the γ value, we design a backward-selection sentinel network mining algorithm (SNMA) for deriving the sentinel network via group sparse Bayesian learning. In order to be practically useful, we further address the issue of scalability in the computation of SNMA, and moreover, extend SNMA to the case of a non-linear dynamical system that could involve complex diffusion mechanisms. We show the effectiveness of SNMA by validating it using both synthetic datasets and five real-world datasets. The experimental results are appealing, which demonstrate that SNMA readily outperforms the state-of-the-art methods.

Although many graph convolutional neural networks (GCNNs) have achieved superior performances in semisupervised node classification, they are designed from either the spatial or spectral perspective, yet without a general theoretical basis. Besides, most of the existing GCNNs methods tend to ignore the ubiquitous noises in the network topology and node content and are thus unable to model these uncertainties. These drawbacks certainly reduce their effectiveness in integrating network topology and node content. To provide a probabilistic perspective to the GCNNs, we model the semisupervised node classification problem as a topology-constrained probabilistic latent space model, probabilistic graph convolutional network (PGCN). By representing the nodes in a more efficient distribution form, the proposed framework can seamlessly integrate the node content and network topology. When specifying the distribution in PGCN to be a Gaussian distribution, the transductive node classification problems can be solved by the general framework and a specific method, called PGCN with the Gaussian distribution representation (PGCN-G), is proposed. To overcome the overfitting problem in covariance estimation and reduce the computational complexity, PGCN-G is further improved to PGCN-G+ by imposing the covariance matrices of all vertices to possess the identical singular vectors. The optimization algorithm based on expectation-maximization indicates that the proposed method can iteratively denoise the network topology and node content with respect to each other. Besides the effectiveness of this top-down framework demonstrated via extensive experiments, it can also be deduced to cover the existing methods, graph convolutional network, graph attention network, and Gaussian mixture model and elaborate their characteristics and relationships by specific derivations.

We propose a general and scalable global optimization framework directly operating on annotated graph data by introducing a Bayesian graph neural network to approximate the expensive-to-evaluate objectives. It prevents the cubical complexity of Gaussian processes and can scale linearly with the number of observations. Its parallelized variant makes it scalable. We provide strict theoretical support on its convergence. Intensive experiments conducted on both artificial and real-world problems, including molecular discovery and urban road network design, demonstrate the effectiveness of the proposed methods compared with the current state of the art.

Stochastic blockmodel (SBM) is a widely used statistical network representation model, with good interpretability, expressiveness, generalization, and flexibility, which has become prevalent and important in the field of network science over the last years. However, learning an optimal SBM for a given network is an NP-hard problem. This results in significant limitations when it comes to applications of SBMs in large-scale networks, because of the significant computational overhead of existing SBM models, as well as their learning methods. Reducing the cost of SBM learning and making it scalable for handling large-scale networks, while maintaining the good theoretical properties of SBM, remains an unresolved problem. In this work, we address this challenging task from a novel perspective of model redefinition. We propose a novel redefined SBM with Poisson distribution and its block-wise learning algorithm that can efficiently analyse large-scale networks. Extensive validation conducted on both artificial and real-world data shows that our proposed method significantly outperforms the state-of-the-art methods in terms of a reasonable trade-off between accuracy and scalability. 1

Recent research on graph embedding has achieved success in various applications. Most graph embedding methods preserve the proximity in a graph into a manifold in an embedding space. We argue an important but neglected problem about this proximity-preserving strategy: Graph topology patterns, while preserved well into an embedding manifold by preserving proximity, may distort in the ambient embedding Euclidean space, and hence to detect them becomes difficult for machine learning models. To address the problem, we propose curvature regularization, to enforce flatness for embedding manifolds, thereby preventing the distortion. We present a novel angle-based sectional curvature, termed ABS curvature, and accordingly three kinds of curvature regularization to induce flat embedding manifolds during graph embedding. We integrate curvature regularization into five popular proximity-preserving embedding methods, and empirical results in two applications show significant improvements on a wide range of open graph datasets.

Finding hidden structural patterns is a critical problem for all types of networks, including signed networks. Among all of the methods for structural analysis of complex network, stochastic blockmodel (SBM) is an important research tool because it is flexible and can generate networks with many different types of structures. However, most existing SBM learning methods for signed networks are unsupervised, leading to poor performance in terms of finding hidden structural patterns, especially when handling noisy and sparse networks. Learning SBM in a semi-supervised way is a promising avenue for overcoming the above difficulty. In this type of model, a small number of labelled nodes and a large number of unlabelled nodes, coupled with their network structures, are simultaneously used to train SBM. We propose a novel semi-supervised signed stochastic blockmodel and its learning algorithm based on variational Bayesian inference, with the goal of discovering both assortative (the nodes connect more densely in same clusters than that in different clusters) and disassortative (the nodes link more sparsely in same clusters than that in different clusters) structures from signed networks. The proposed model is validated through a number of experiments wherein it compared with the state-of-the-art methods using both synthetic and real-world data. The carefully designed tests, allowing to account for different scenarios, show our method outperforms other approaches existing in this space. It is especially relevant in the case of noisy and sparse networks as they constitute the majority of the real-world networks.

Attributed graphs, which contain rich contextual features beyond just network structure, are ubiquitous and have been observed to benefit various network analytics applications. Graph structure optimization, aiming to find the optimal graphs in terms of some specific measures, has become an effective computational tool in complex network analysis. However, traditional model-free methods suffer from the expensive computational cost of evaluating graphs; existing vectorial Bayesian optimization methods cannot be directly applied to attributed graphs and have the scalability issue due to the use of Gaussian processes (GPs). To bridge the gap, in this paper, we propose a novel scalable Deep Graph Bayesian Optimization (DGBO) method on attributed graphs. The proposed DGBO prevents the cubical complexity of the GPs by adopting a deep graph neural network to surrogate black-box functions, and can scale linearly with the number of observations. Intensive experiments are conducted on both artificial and real-world problems, including molecular discovery and urban road network design, and demonstrate the effectiveness of the DGBO compared with the state-of-the-art.

In the past few years, semi-supervised node classification in attributed network has been developed rapidly. Inspired by the success of deep learning, researchers adopt the convolutional neural network to develop the Graph Convolutional Networks (GCN), and they have achieved surprising classification accuracy by considering the topological information and employing the fully connected network (FCN). However, the given network topology may also induce a performance degradation if it is directly employed in classification, because it may possess high sparsity and certain noises. Besides, the lack of learnable filters in GCN also limits the performance. In this paper, we propose a novel Topology Optimization based Graph Convolutional Networks (TO-GCN) to fully utilize the potential information by jointly refining the network topology and learning the parameters of the FCN. According to our derivations, TO-GCN is more flexible than GCN, in which the filters are fixed and only the classifier can be updated during the learning process. Extensive experiments on real attributed networks demonstrate the superiority of the proposed TO-GCN against the state-of-the-art approaches.

Image recognition aims to identify objects, places, people, or other targeted items in a given image, and has a wide range of social applications such as natural disasters recognition, plant disease detection, and traffic jam detection. Currently state-of-the-art methods of image recognition are based on deep learning and remain a common pattern in designing and using convolutional neural networks (CNNs). However, designing CNNs is extremely time intensive and requires an expert. Neural architecture search (NAS) can solve this problem by automatically identifying architectures of CNNs that are superior to hand-designed ones. Recently BO has been applied to neural architecture search and shows better performance than pure evolutionary strategies. All these methods adopt Gaussian processes (GPs) as surrogate function, with the handcraft similarity metrics as input. In this work, we propose a Bayesian graph neural network as a new surrogate, which can automatically extract features from deep neural architectures, and use such learned features to fit and characterize black-box objectives and their uncertainty. Based on the new surrogate, we then develop a graph Bayesian optimization framework to address the challenging task of deep neural architecture search. Experiment results show our method significantly outperforms the comparative methods on benchmark tasks.

Predicting epidemic dynamics is of great value in understanding and controlling diffusion processes, such as infectious disease spread and information propagation. This task is intractable, especially when surveillance resources are very limited. To address the challenge, we study the problem of active surveillance, i.e., how to identify a small portion of system components as sentinels to effect monitoring, such that the epidemic dynamics of an entire system can be readily predicted from the partial data collected by such sentinels. We propose a novel measure, the gamma value, to identify the sentinels by modeling a sentinel network with row sparsity structure. We design a flexible group sparse Bayesian learning algorithm to mine the sentinel network suitable for handling both linear and non-linear dynamical systems by using the expectation maximization method and variational approximation. The efficacy of the proposed algorithm is theoretically analyzed and empirically validated using both synthetic and real-world data.

Traditional supervised text classifiers require a large number of manually labeled documents, which are often expensive to obtain. Recently, dataless text classification has attracted more attention, since it only requires very few seed words of categories that are much cheaper. In this paper, we develop a pseudo-label based dataless Naive Bayes (PL-DNB) classifier with seed words. We initialize pseudo-labels for each document using seed word occurrences, and employ the expectation maximization algorithm to train PL-DNB in a semi-supervised manner. The pseudo-labels are iteratively updated using a mixture of seed word occurrences and estimations of label posteriors. To avoid noisy pseudo-labels, we also consider the information of nearest neighboring documents in the pseudo-label update step, i.e., preserving local neighborhood structure of documents. We empirically show that PL-DNB outperforms traditional dataless text classification algorithms with seed words. Especially, PL-DNB performs well on the imbalanced dataset.

Recommender systems are used to accurately and actively provide users with potentially interesting information or services. Collaborative filtering is a widely adopted approach to recommendation, but sparse data and cold-start users are often barriers to providing high quality recommendations. To address such issues, we propose a novel method that works to improve the performance of collaborative filtering recommendations by integrating sparse rating data given by users and sparse social trust network among these same users. This is a model-based method that adopts matrix factorization technique that maps users into low-dimensional latent feature spaces in terms of their trust relationship, and aims to more accurately reflect the users reciprocal influence on the formation of their own opinions and to learn better preferential patterns of users for high-quality recommendations. We use four large-scale datasets to show that the proposed method performs much better, especially for cold start users, than state-of-the-art recommendation algorithms for social collaborative filtering based on trust.

During an epidemic, the spatial, temporal and demographic patterns of disease transmission are determined by multiple factors. In addition to the physiological properties of the pathogens and hosts, the social contact of the host population, which characterizes the reciprocal exposures of individuals to infection according to their demographic structure and various social activities, are also pivotal to understanding and predicting the prevalence of infectious diseases. How social contact is measured will affect the extent to which we can forecast the dynamics of infections in the real world. Most current work focuses on modeling the spatial patterns of static social contact. In this work, we use a novel perspective to address the problem of how to characterize and measure dynamic social contact during an epidemic. We propose an epidemic-model-based tensor deconvolution framework in which the spatiotemporal patterns of social contact are represented by the factors of the tensors. These factors can be discovered using a tensor deconvolution procedure with the integration of epidemic models based on rich types of data, mainly heterogeneous outbreak surveillance data, socio-demographic census data and physiological data from medical reports. Using reproduction models that include SIR/SIS/SEIR/SEIS models as case studies, the efficacy and applications of the proposed framework are theoretically analyzed, empirically validated and demonstrated through a set of rigorous experiments using both synthetic and real-world data.

Signed networks with positive and negative links attract considerable interest in their studying since they contain more information than unsigned networks. Community detection and sign (or attitude) prediction are still primary challenges, as the fundamental problems of signed network analysis. For this, a generative Bayesian approach is presented wherein 1) a signed stochastic blockmodel is proposed to characterize the community structure in the context of signed networks, by explicit formulating the distributions of the density and frustration of signed links from a stochastic perspective, and 2) a model learning algorithm is advanced by theoretical deriving a variational Bayes EM for the parameter estimation and variation-based approximate evidence for the model selection. The comparison of the above approach with the state-of-the-art methods on synthetic and real-world networks, shows its advantage in the community detection and sign prediction for the exploratory networks.

Stochastic blockmodel (SBM) enables us to decompose and analyze an exploratory network without a priori knowledge about its intrinsic structure. However, the task of effectively and efficiently learning a SBM from a large-scale network is still challenging due to the high computational cost of its model selection and parameter estimation. To address this issue, we present a novel SBM learning algorithm referred to as BLOS (BLOckwise Sbm learning). Distinct from the literature, the model selection and parameter estimation of SBM are concurrently, rather than alternately, executed in BLOS by embedding the minimum message length criterion into a block-wise EM algorithm, which greatly reduces the time complexity of SBM learning without losing learning accuracy and modeling flexibility. Its effectiveness and efficiency have been tested through rigorous comparisons with the state-of-the-art methods on both synthetic and real-world networks.

There has been an increasing interest in exploring signed networks with positive and negative links in that they contain more information than unsigned networks. As fundamental problems of signed network analysis, community detection and sign (or attitude) prediction are still primary challenges. To address them, we propose a generative Bayesian approach, in which 1) a signed stochastic blockmodel is proposed to characterize the community structure in context of signed networks, by means of explicitly formulating the distributions of both density and frustration of signed links from a stochastic perspective, and 2) a model learning algorithm is proposed by theoretically deriving a variational Bayes EM for parameter estimation and a variation based approximate evidence for model selection. Through the comparisons with state-of-the-art methods on synthetic and real-world networks, the proposed approach shows its superiority in both community detection and sign prediction for exploratory networks.

Active surveillance is a desirable way to prevent the spread of infectious diseases in that it aims to timely discover individual incidences through an active searching for patients. However, in practice active surveillance is difficult to implement especially when monitoring space is large but available resources are limited. Therefore, it is extremely important for public health authorities to know how to distribute their very sparse resources to high-priority regions so as to maximize the outcomes of active surveillance. In this paper, we raise the problem of active surveillance planning and provide an effective method to address it via modeling and mining spatiotemporal patterns of infection risks from heterogeneous data sources. Taking malaria as an example, we perform an empirical study on real-world data to validate our method and provide our new findings.

During an epidemic, the spatial, temporal and demographical patterns of disease transmission are determined by multiple factors. Besides the physiological properties of pathogenes and hosts, the social contacts of host population, which characterize individuals' reciprocal exposures of infection in view of demographical structures and various social activities, are also pivotal to understand and further predict the prevalence of infectious diseases. The means of measuring social contacts will dominate the extent how precisely we can forecast the dynamics of infections in the real world. Most current works focus their efforts on modeling the spatial patterns of static social contacts. In this work, we address the problem on how to characterize and measure dynamical social contacts during an epidemic from a novel perspective. We propose an epidemic-model-based tensor deconvolution framework to address this issue, in which the spatiotemporal patterns of social contacts are represented by the factors of tensors, which can be discovered by a tensor deconvolution procedure with an integration of epidemic models from rich types of data, mainly including heterogeneous outbreak surveillance, social-demographic census and physiological data from medical reports. Taking SIR model as a case study, the efficacy of the proposed method is theoretically analyzed and empirically validated through a set of rigorous experiments on both synthetic and real-world data.

The real-world continuous double auction (CDA) market is a dynamic environment. However, most of the existing agent bidding strategies are simply designed for static markets. A new detecting method for bidding strategy is necessary for more practical

Network communities refer to groups of vertices within which their connecting links are dense but between which they are sparse. A network community mining problem (or NCMP for short) is concerned with the problem of finding all such communities from a given network. A wide variety of applications can be formulated as NCMPs, ranging from social and/or biological network analysis to web mining and searching. So far, many algorithms addressing NCMPs have been developed and most of them fall into the categories of either optimization based or heuristic methods. Distinct from the existing studies, the work presented in this paper explores the notion of network communities and their properties based on the dynamics of a stochastic model naturally introduced. In the paper, a relationship between the hierarchical community structure of a network and the local mixing properties of such a stochastic model has been established with the large-deviation theory. Topological information regarding to the community structures hidden in networks can be inferred from their spectral signatures. Based on the above-mentioned relationship, this work proposes a general framework for characterizing, analyzing, and mining network communities. Utilizing the two basic properties of metastability, i.e., being locally uniform and temporarily fixed, an efficient implementation of the framework, called the LM algorithm, has been developed that can scalably mine communities hidden in large-scale networks. The effectiveness and efficiency of the LM algorithm have been theoretically analyzed as well as experimentally validated.

Complex network theory provides a means for modeling and analyzing complex systems that consist of multiple and interdependent components. Among the studies on complex networks, structural analysis is of fundamental importance as it presents a natural route to understanding the dynamics, as well as to synthesizing or optimizing the functions, of networks. A wide spectrum of structural patterns of networks has been reported in the past decade, such as communities, multipartites, bipartite, hubs, authorities, outliers, and bow ties, among others. In this paper, we are interested in tackling the challenging task of characterizing and extracting multiplex patterns (multiple patterns as mentioned previously coexisting in the same networks in a complicated manner), which so far has not been explicitly and adequately addressed in the literature. Our work shows that such multiplex patterns can be well characterized as well as effectively extracted by means of a granular stochastic blockmodel, together with a set of related algorithms proposed here based on some machine learning and statistical inference ideas. These models and algorithms enable us to further explore complex networks from a novel perspective.

The ability to discover community structures from explorative networks is useful for many applications. Most of the existing methods with regard to community mining are specifically designed for assortative networks, and some of them could be applied to address disassortative networks by means of intentionally modifying the objectives to be optimized. However, the types of the explorative networks are unknown beforehand. Consequently, it is difficult to determine what specific algorithms should be used to mine appropriate structures from exploratory networks. To address this issue, a novel concept, generalized community structure, has been proposed with the attempt to unify the two distinct counterparts in both types of networks. Furthermore, based on the proposed random network ensemble model, a generalized community mining algorithm, so called G-NCMA, has been proposed, which is promisingly suitable for both types of networks. Its performance has been rigorously tested, validated and compared with other related algorithms against real-world networks as well as synthetic networks. Experimental results show the G-NCMA algorithm is able to detect communities, without any prior, from explorative networks with a good accuracy.

The ability to discover patterns of networks is fundamental for structural analysis applied to them. Many ubiquitous patterns demonstrated by real-world networks have been discovered, and corresponding tools for finding them also have been developed. Although existing works have greatly improved our understanding on networks, it is still challenging to precisely model and predict their behaviors mainly because their non-trivial structures usually consists of many fold coexisting patterns which cannot be appropriately and totally uncovered by a single tool exclusively designed for pre-defined ones. In this work, we take an effort to address this issue by introducing a parameter-free algorithm aiming to discover such patterns hidden in an explorative network.

Research has shown that many social networks come into being hierarchically based on some basic building blocks called communities, within which the social interactions are very intensive, but between which they are very weak. Network community mining algorithms aim at efficiently and effectively discovering all such communities from a given network. Many related methods have been proposed and applied to different areas including social network analysis, gene network analysis and web clustering engine. Most of the existing methods for mining communities are centralized. In this paper, we present a multi-agent based decentralized algorithm, in which a group of autonomous agents work together to mine a network through a proposed self-aggregation and self-organization mechanism. Thanks to its decentralized feature, our method is potentially suitable for dealing with distributed networks, whose global structures are hard to obtain due to their geographical distributions, decentralized controls or huge sizes. The effectiveness of our method has been tested against different benchmark networks.

Recently, the sizes of networks are always very huge, and they take on distributed nature. Aiming at this kind of network clustering problem, in the sight of local view, this paper proposes a fast network clustering algorithm in which each node is regarded as an agent, and each agent tries to maximize its local function in order to optimize network modularity defined by function Q, rather than optimize function Q from the global view as traditional methods. Both the efficiency and effectiveness of this algorithm are tested against computer-generated and real-world networks. Experimental result shows that this algorithm not only has the ability of clustering large-scale networks, but also can attain very good clustering quality compared with the existing algorithms. Furthermore, the parameters of this algorithm are analyzed.

The term of social network communities refers to groups of individuals within which social interactions are intense and between which they are weak. A social network community mining problem (SNCMP) can be stated as the problem of finding all such communities from a given social network. A wide variety of applications can be formulated into SNCMPs, ranging from Web intelligence to social intelligence. So far, many algorithms addressing the SNCMP have been developed; most of them are either optimization or heuristic based methods. Different from all existing work, this paper explores the notion of a social network community and its intrinsic properties, drawing on the dynamics of a stochastic model naturally introduced. In particular, it uncovers an interesting connection between the hierarchical community structure of a network and the metastability of a Markov process constructed upon it. A lot of critical topological information regarding to communities hidden in networks can be inferred from the derived spectral signatures of such networks, without actually clustering them with any particular algorithms. Based upon the above connection, we can obtain a frameworkfor characterizing and analyzing social network communities.

One of the central problems in studying and understanding complex networks, such as online social networks or World Wide Web, is to discover hidden, either physically (e.g., interactions or hyperlinks) or logically (e.g., profiles or semantics) well-defined topological structures. From a practical point of view, a good example of such structures would be so-called network communities. Earlier studies have introduced various formulations as well as methods for the problem of identifying or extracting communities. While each of them has pros and cons as far as the effectiveness and efficiency are concerned, almost none of them has explicitly dealt with the potential relationship between the global topological property of a network and the local property of individual nodes. In order to study this problem, this paper presents a new algorithm, called ICS, which aims to discover natural network communities by inferring from the local information of nodes inherently hidden in networks based on a new centrality, that is, clustering centrality, which is a generalization of eigenvector centrality. As compared with existing methods, our method runs efficiently with a good clustering performance. Additionally, it is insensitive to its built-in parameters and prior knowledge.

Many complex systems in the real world can be modeled as signed social networks that contain both positive and negative relations. Algorithms for mining social networks have been developed in the past, however most of them were designed primarily for networks containing only positive relations and thus not suitable for signed networks. In this work, we propose a new algorithm, called FEC, to mine signed social networks so that both positive within-group relations and negative between-group relations are dense. FEC considers both the sign and the density of relations as the clustering attributes, making itself effective for not only signed networks but also conventional social networks including only positive relations. Also, FEC adopts an agent-based heuristic that makes the algorithm efficient (in linear time with respect to the size of a network) and capable of giving nearly optimal solutions. FEC depends on only one parameter whose value can easily be set, and requires no prior knowledge on hidden community structures. The effectiveness and efficacy of FEC have been demonstrated through a set of rigorous experiments involving both benchmark and randomly-generated signed networks.