The evolution of Industry 5.0 introduces complex, temporally irregular network behaviors, challenging intrusion detection in heterogeneous industrial environments. Existing Intrusion Detection Systems (IDSs) often struggle with asynchronous traffic and lack interpretability. To overcome these limitations, we propose AegisNet, a novel sequence-based IDS that reformulates intrusion detection as a sequence classification task on fixed-length windows ofTconsecutive observations (sequence length), augmented with explicit time-delta features to capture irregular traffic intervals. AegisNet introduces a novel Temporal-Aware and Self-Adaptive Multi-Head Attention (TA-SAMHA) mechanism. It introduces learnable temporal bias projections to directly encode irregular time intervals into the attention process, enabling the system to capture fine-grained temporal dependencies and evolving behaviors. It also employs a Dual-Gate Residual Encoder (DGRE) for bidirectional sequence modeling with dual gating, enabling enriched contextual representations and stable learning across deep temporal layers. Additionally, we introduce the IDS-specific adaptation of Temporal SHAP (T-SHAP), enabling time-resolved feature attributions that expose how influential factors shift during sequential intrusion detection, offering a new dimension of temporally aware explainability for evolving attack patterns. Experimental results demonstrate that AegisNet achieves a detection accuracy of 99.63% on the CICIoV2024 dataset and 99.13% on the CICIoT2023 dataset, demonstrating its effectiveness for securing Industrial environments.

The emergence of Blockchain-enabled Internet of Vehicles stands as a critical method for secure data exchange between vehicles and traffic infrastructures. Yet, significant challenges persist, particularly in terms of limited fault tolerance and constraints associated with storage and computational resources. In this study, we introduce the Stateless-based Lightweight Sharding Mechanism (SLIM) for enhanced secure data exchange. This mechanism is built on three key strategies: 1) SLIM incorporates a robust sharding protocol which includes identity establishment and verification, shard formation, and intra-shard consensus to mitigate the impact of Byzantine nodes within each shard. This protocol is designed to reduce the influence of malicious nodes in each shard, thereby improving fault tolerance and security. 2) To address storage challenges, we employ a stateless verification algorithm to allow service nodes in the Internet of Vehicles to authenticate vehicular data transactions without the need to store the entire historical blockchain data. 3) SLIM also integrates a Stackelberg game model for efficient resource distribution in collaborative networks. By offloading the mining process and data storage to the cloud and allowing roadside units to adjust their storage and computing strategies through the game model, the computing resource requirement of roadside units is thus substantially reduced, leading to optimized revenue generation. The security effectiveness of SLIM is theoretically proven. The simulations and experimental results demonstrate that SLIM’s block size is significantly smaller (66.9 times less) than Ethereum’s, and its block verification time is 67% faster compared to conventional stateless blockchains.

Abstract To monitor clients potentially bypassing position limits, business units employ the disjoint sets principle to identify potential client correlations based on account profiles. The key challenge lies in computing disjoint sets for large-scale topological graphs (with millions of nodes and billions of edges) in a short response time. In this article, we propose a multi-DAG indexing algorithm, namely the Hierarchical Isomerism Distributed Equivalent (HIDE) union find. First, in large-scale topological graphs, we utilize two new topological structures: the equivalent sub-Directed Acyclic Graph (sub-DAG) and the hierarchical isomerism topological graph, to reduce the number of edges and nodes in the multi-DAG indexing merging process. Then, our HIDE union find is proposed to achieve computable splitting across temporal and spatial spans. HIDE union find is theoretically proven to ensure correctness and universality. Experimental validation demonstrates that HIDE union find outperforms previous methods in large-scale topological graphs. The results indicate that HIDE union find achieves response times 100 to 200 times faster than those of the current leading methods.

Temporal link prediction is one of the most important tasks for predicting time-varying links by capturing dynamics within complex networks. However, it suffers from difficulties such as vulnerability to adversarial attacks and inadaptation to distinct evolutionary patterns. In this article, we propose a robust temporal link prediction architecture via stable gated models with reinforcement learning (SAGE-RL) consisting of a state encoding network (SEN) and a self-adaptive policy network (SPN). The former is utilized to capture network dynamics, while the latter helps the former adapt to distinct evolutionary patterns across various time periods. Within the SEN, a novel stable gate is introduced to ensure multiple spatiotemporal dependency paths and defend against adversarial attacks. An SPN is proposed to select different SEN instances by approximating the optimal action function, thereby adapting to various evolutionary patterns to learn the robust temporal and structural features from dynamic complex networks. It is proven that SAGE-LR with integral Lipschitz graph convolution is stable to relative perturbations in dynamic complex networks. With the aid of extensive experiments on five real-world graph benchmarks, SAGE-LR is shown to substantially outperform current state-of-the-art approaches in terms of precision and stability of temporal link prediction and ability to successfully defend against various attacks. We also implement the temporal link prediction in shipping transaction networks, which forecast effectively its potential transaction risks.

Abstract As a structural topological index, the number of subtrees has great significance for the analysis and design of hybrid locally reliable networks. In this paper, with generating function and introducing a novel two-forest dual transformation technique, we solve the subtree enumerating problems of two representatives of the self-similar networks, such as the hierarchical lattice and $(u,v)$-flower networks. Moreover, by means of the circle weight transfer technique, two linear time algorithms of computing the subtree generation functions of these two families of networks are also proposed. The subtree density of two special cases for these self-similar networks is briefly discussed as an application.

Recommendation systems have become an indispensable part of daily life. Social recommendation systems, which utilize social relationships and past behaviors to infer users’ preferences, have gained popularity in recent years. Exploring the inherent characteristics implied by higher-order relationships offers a new approach to social recommendation. However, it is challenging due to sparse social networks, influence heterogeneity, and noisy feedback. In this article, we propose a Heterogeneous Meta-path Graph Learning model for Higher-order Social Recommendation (HEAL). Within HEAL, we introduce a heterogeneous graph in social recommendation and utilize a meta-path-guided random walk to generate higher-order relationships. By encoding higher-order structures and semantics along different meta-graphs, HEAL can mitigate the limitation of data sparsity. Moreover, HEAL exploits aspect-aware and semantic-aware attentions to adaptively propagate and aggregate useful features from different meta-neighbors and higher-order relations. These attention-based aggregation layers allow HEAL to suppress the heterogeneity of social influences. Furthermore, HEAL adopts contrastive learning as a supplemental task to the recommendation task by maximizing the consistency between the self-discriminating objectives. This auxiliary task enables the model to learn more differentiated representations, further reducing its sensitivity to noisy feedback. We evaluate the performance of HEAL through extensive experiments on public datasets. The results demonstrate that leveraging higher-order relations can enhance the quality of social recommendations by better capturing the complexity and diversity of users’ preferences and interactions.

A graph neural network (GNN) is a powerful architecture for semi-supervised learning (SSL). However, the data-driven mode of GNNs raises some challenging problems. In particular, these models suffer from the limitations of incomplete attribute learning, insufficient structure capture, and the inability to distinguish between node attribute and graph structure, especially on label-scarce or attribute-missing data. In this article, we propose a novel framework, called graph coneighbor neural network (GCoNN), for node classification. It is composed of two modules: GCoNN$_{\Gamma}$and GCoNN$_{\mathop{\Gamma}\limits^{\circ}}$. GCoNN$_{\Gamma}$is trained to establish the fundamental prototype for attribute learning on labeled data, while GCoNN$_{\mathring{\Gamma}}$learns neighbor dependence on transductive data through pseudolabels generated by GCoNN$_{\Gamma}$. Next, GCoNN$_{\Gamma}$is retrained to improve integration of node attribute and neighbor structure through feedback from GCoNN$_{\mathring{\Gamma}}$. GCoNN tends to convergence iteratively using such an approach. From a theoretical perspective, we analyze this iteration process from a generalized expectation–maximization (GEM) framework perspective which optimizes an evidence lower bound (ELBO) by amortized variational inference. Empirical evidence demonstrates that the state-of-the-art performance of the proposed approach outperforms other methods. We also apply GCoNN to brain functional networks, the results of which reveal response features across the brain which are physiologically plausible with respect to known language and visual functions.

Predicting the future trajectories of pedestrians is of increasing importance for many applications such as autonomous driving and social robots. Nevertheless, current trajectory prediction models suffer from limitations such as lack of diversity in candidate trajectories, poor accuracy, and instability. In this paper, we propose a novel Sequence Entropy Energy-based Model named SEEM, which consists of a generator network and an energy network. Within SEEM we optimize the sequence entropy by taking advantage of the local variational inference of f-divergence estimation to maximize the mutual information across the generator in order to cover all modes of the trajectory distribution, thereby ensuring SEEM achieves full diversity in candidate trajectory generation. Then, we introduce a probability distribution clipping mechanism to draw samples towards regions of high probability in the trajectory latent space, while our energy network determines which trajectory is most representative of the ground truth. This dual approach is our so-called all-then-one strategy. Finally, a zero-centered potential energy regularization is proposed to ensure stability and convergence of the training process. Through experiments on both synthetic and public benchmark datasets, SEEM is shown to substantially outperform the current state-of-the-art approaches in terms of diversity, accuracy and stability of pedestrian trajectory prediction.

Temporal link prediction (TLP) is among the most important graph learning tasks, capable of predicting dynamic, time-varying links within networks. The key problem of TLP is how to explore potential link-evolving tendency from the increasing number of links over time. There exist three major challenges toward solving this problem: temporal nonlinear sparsity, weak serial correlation, and discontinuous structural dynamics. In this article, we propose a novel transfer learning model, called DNformer, to predict temporal link sequence in dynamic networks. The structural dynamic evolution is sequenced into consecutive links one by one over time to inhibit temporal nonlinear sparsity. The self-attention of the model is used to capture the serial correlation between the input and output link sequences. Moreover, our structural encoding is designed to obtain changing structures from the consecutive links and to learn the mapping between link sequences. This structural encoding consists of two parts: the node clustering encoding of each link and the link similarity encoding between links. These encodings enable the model to perceive the importance and correlation of links. Furthermore, we introduce a measurement of structural similarity in the loss function for the structural differences of link sequences. The experimental results demonstrate that our model outperforms other state-of-the-art TLP methods such as Transformer, TGAT, and EvolveGCN. It achieves the three highest AUC and four highest precision scores in five different representative dynamic networks problems.

Abstract The subtree number index of a graph, defined as the number of subtrees, attracts much attention recently. Finding a proper algorithm to compute this index is an important but difficult problem for a general graph. Even for unicyclic and bicyclic graphs, it is not completely trivial, though it can be figured out by try and error. However, it is complicated for tricyclic graphs. This paper proposes path contraction carrying weights (PCCWs) algorithms to compute the subtree number index for the nontrivial case of bicyclic graphs and all 15 cases of tricyclic graphs, based on three techniques: PCCWs, generating function and structural decomposition. Our approach provides a foundation and useful methods to compute subtree number index for graphs with more complicated cycle structures and can be applied to investigate the novel structural property of some important nanomaterials such as the pentagonal carbon nanocone.

Detecting self-organized coalitions from functional networks is one of the most important ways to uncover functional mechanisms in the brain. Determining these raises well-known technical challenges in terms of scale imbalance, outliers and hard-examples. In this article, we propose a novel self-adaptive skeleton approach to detect coalitions through an approximation method based on probabilistic mixture models. The nodes in the networks are characterized in terms of robust k -order complete subgraphs ( k -clique ) as essential substructures. The k -clique enumeration algorithm quickly enumerates all k -cliques in a parallel manner for a given network. Then, the cliques, from max -clique down to min -clique, of each order k , are hierarchically embedded into a probabilistic mixture model. They are self-adapted to the corresponding structure density of coalitions in the brain functional networks through different order k . All the cliques are merged and evolved into robust skeletons to sustain each unbalanced coalition by eliminating outliers and separating overlaps. We call this the k -CLIque Merging Evolution (CLIME) algorithm. The experimental results illustrate that the proposed approaches are robust to density variation and coalition mixture and can enable the effective detection of coalitions from real brain functional networks. There exist potential cognitive functional relations between the regions of interest in the coalitions revealed by our methods, which suggests the approach can be usefully applied in neuroscientific studies.

In this article, we propose a framework for crowd behavior prediction in complicated scenarios. The fundamental framework is designed using the standard encoder-decoder scheme, which is built upon the long short-term memory module to capture the temporal evolution of crowd behaviors. To model interactions among humans and environments, we embed both the social and the physical attention mechanisms into the long short-term memory. The social attention component can model the interactions among different pedestrians, whereas the physical attention component helps to understand the spatial configurations of the scene. Since pedestrians’ behaviors demonstrate multi-modal properties, we use the generative model to produce multiple acceptable future paths. The proposed framework not only predicts an individual’s trajectory accurately but also forecasts the ongoing group behaviors by leveraging on the coherent filtering approach. Experiments are carried out on the standard crowd benchmarks (namely, the ETH, the UCY, the CUHK crowd, and the CrowdFlow datasets), which demonstrate that the proposed framework is effective in forecasting crowd behaviors in complex scenarios.

Mining the relationship structures among the investors plays a vital role in promoting economic development as well as preventing financial risks, especially in the context of big data. This article proposes fast networking approaches from investment big data to explore three underlying structures, namely, investment pedigrees, investment groups, and structural holes. Inspired by disjoint sets and path compression, we first present a pedigree classification algorithm to identify investment pedigrees. Second, through introducing a pruning strategy and a data structure termed as “2-tuple list,” we develop a novel linear-time structure mining algorithm in network (SMAN) for investigating investment groups and structural holes from the investment pedigree. Finally, we show that our SMAN has higher clustering accuracy and efficiency than other existing algorithms on a variety of real-world tasks in terms of normalized mutual information (NMI) values. Our method is particularly well suited for mining the underlying structures from investment big data.

Missing value (MV) imputation is a critical preprocessing means for data mining. Nevertheless, existing MV imputation methods are mostly designed for batch processing, and thus are not applicable to streaming data, especially those with poor quality. In this article, we propose a framework, called Real-time and Error-tolerant Missing vAlue ImputatioN (REMAIN), to impute MVs in poor-quality streaming data. Instead of imputing MVs based on all the observed data, REMAIN first initializes the MV imputation model based on a-RANSAC which is capable of detecting and rejecting anomalies in an efficient manner, and then incrementally updates the model parameters upon the arrival of new data to support real-time MV imputation. As the correlations among attributes of the data may change over time in unforseenable ways, we devise a deterioration detection mechanism to capture the deterioration of the imputation model to further improve the imputation accuracy. Finally, we conduct an extensive evaluation on the proposed algorithms using real-world and synthetic datasets. Experimental results demonstrate that REMAIN achieves significantly higher imputation accuracy over existing solutions. Meanwhile, REMAIN improves up to one order of magnitude in time cost compared with existing approaches.

A comprehensive knowledge system reveals the intangible insights hidden in an information system by integrating information from multiple data sources in a synthetical manner. In this paper, we present a variable precision reduction theory, underpinned by two new concepts: 1) distribution tables and 2) genealogical binary trees. Sufficient and necessary conditions to extract comprehensive knowledge from a given information system are also presented and proven. A complete variable precision reduction algorithm is proposed, in which we introduce four important strategies, namely, distribution table abstracting, attribute rank dynamic updating, hierarchical binary classifying, and genealogical tree pruning. The completeness of our algorithm is proven theoretically and its superiority to existing methods for obtaining complete reducts is demonstrated experimentally. Finally, having obtaining the complete reduct set, we demonstrate how the relationships between the complete reduct set and the comprehensive knowledge system can be visualized in a double-layer lattice structure using Hasse diagrams.

Sequence transfer learning is of interest in both academia and industry with the emergence of numerous new text domains from Twitter and other social media tools. In this paper, we put forward the data-sensitive granularity for transfer learning, and then, a novel substructural regularization transfer learning model (STLM) is proposed to preserve target domain features at substructural granularity in the light of the condition of labeled data set size. Our model is underpinned by hidden Markov model and regularization theory, where the substructural representation can be integrated as a penalty after measuring the dissimilarity of substructures between target domain and STLM with relative entropy. STLM can achieve the competing goals of preserving the target domain substructure and utilizing the observations from both the target and source domains simultaneously. The estimation of STLM is very efficient since an analytical solution can be derived as a necessary and sufficient condition. The relative usability of substructures to act as regularization parameters and the time complexity of STLM are also analyzed and discussed. Comprehensive experiments of part-of-speech tagging with both Brown and Twitter corpora fully justify that our model can make improvements on all the combinations of source and target domains.

As one of the counting-based topological indices, the number of subtrees and its variations has received much attention in recent years. In this paper, using generating functions, we investigate and derive formulas for this index of hexagonal and phenylene chains. We also present graph-theoretical algorithms for enumerating subtrees of these two chains. Extremal values and graphs with respect to the subtree number among all hexagonal and phenylene chains with n hexagons are also determined. As an application, we briefly examine the subtree densities of these two chains.

The central time series crystallizes the common patterns of the set it represents. In this paper, we propose a global constrained degree-pruning dynamic programming (g(dp)2) approach to obtain the central time series through minimizing dynamic time warping (DTW) distance between two time series. The DTW matching path theory with global constraints is proved theoretically for our degree-pruning strategy, which is helpful to reduce the time complexity and computational cost. Our approach can achieve the optimal solution between two time series. An approximate method to the central time series of multiple time series [called as m_g(dp)2] is presented based on DTW barycenter averaging and our g(dp)2 approach by considering hierarchically merging strategy. As illustrated by the experimental results, our approaches provide better within-group sum of squares and robustness than other relevant algorithms.

This paper develops an adaptive robust output feedback control scheme for dynamically positioned ships with unavailable velocities and unknown dynamic parameters in an unknown time-variant disturbance environment. The controller is designed by incorporating the high-gain observer and radial basis function (RBF) neural networks in vectorial backstepping method. The high-gain observer provides the estimations of the ship position and heading as well as velocities. The RBF neural networks are employed to compensate for the uncertainties of ship dynamics. The adaptive laws incorporating a leakage term are designed to estimate the weights of RBF neural networks and the bounds of unknown time-variant environmental disturbances. In contrast to the existing results of dynamic positioning (DP) controllers, the proposed control scheme relies only on the ship position and heading measurements and does not require a priori knowledge of the ship dynamics and external disturbances. By means of Lyapunov functions, it is theoretically proved that our output feedback controller can control a ship's position and heading to the arbitrarily small neighborhood of the desired target values while guaranteeing that all signals in the closed-loop DP control system are uniformly ultimately bounded. Finally, simulations involving two ships are carried out, and simulation results demonstrate the effectiveness of the proposed control scheme.

Gaze tracking is crucial for studying driver's attention, detecting fatigue, and improving driver assistance systems, but it is difficult in natural driving environments due to nonuniform and highly variable illumination and large head movements. Traditional calibrations that require subjects to follow calibrators are very cumbersome to be implemented in daily driving situations. A new automatic calibration method, based on a single camera for determining the head orientation and which utilizes the side mirrors, the rear-view mirror, the instrument board, and different zones in the windshield as calibration points, is presented in this paper. Supported by a self-learning algorithm, the system tracks the head and categorizes the head pose in 12 gaze zones based on facial features. The particle filter is used to estimate the head pose to obtain an accurate gaze zone by updating the calibration parameters. Experimental results show that, after several hours of driving, the automatic calibration method without driver's corporation can achieve the same accuracy as a manual calibration method. The mean error of estimated eye gazes was less than 5°in day and night driving.

Context-aware web services have been attracting significant attention as an important approach for improving the usability of web services. In this paper, we explore a novel approach to model dynamic behaviors of interacting context-aware web services, aiming to effectively process and take advantage of contexts and realize behavior adaptation of web services and further to facilitate the development of context-aware application of web services. We present an interaction model of context-aware web services based on context-aware process network (CAPN), which is a data-flow and channel-based model of cooperative computation. The CAPN is extended to context-aware web service network by introducing a kind of sensor processes, which is used to catch contextual data from external environment. Through modeling the register link's behaviors, we present how a web service can respond to its context changes dynamically. The formal behavior semantics of our model is described by calculus of communicating systems process algebra. The behavior adaptation and context awareness in our model are discussed. An eXtensible Markup Language-formatted service behavior description language named BML4WS is designed to describe behaviors and behavior adaptation of interacting context-aware web services. Finally, an application case is demonstrated to illustrate the proposed model how to adapt context changes and describe service behaviors and their changes.

This paper presents a swarm rough approach to analyze the combination factors of violent crime. The approach discovers the feature combinations in an efficient way to observe the change of rough set positive region as the fuzzy swarm proceed throughout the search space. We evaluated the performance of our approach using the violent factor datasets and the corresponding computational experiments are discussed. Empirical results indicate that our approach is ideal for all the considered problems and the fuzzy swarm optimization technique outperforms dynamic reducts (DR) approache by obtaining multiple reductions for the combination factor datasets.

Hidden Markov models (HMM) have been widely used in natural language processing (NLP), especially in syntactic level applications, which appears naturally as short-range-dependent sequence recognition problems. But the structure of HMM limits the usage of global knowledge including the sentiment analysis of the text, which has become an increasingly popular research topic in NLP now. In this paper, we propose a novel treatment of HMM model to use the result of sentimental subjectivity analysis in syntactic level task, i.e. part-of-speech (POS) tagging. The subjectivity information is introduced as a pre-classification procedure into the interval-type HMM. The subjectivity degree of the testing sentence is used as a combination factor to choose an appropriate value from the interval. Experiments results on public tagging data sets shows that the proposed approach enhanced the performance of POS tagging.

Risk assessment is often done by human experts, because there is no exact and mathematical solution to the problem. Usually the human reasoning and perception process cannot be expressed precisely. This paper propose a light weight risk assessment system based on an Hierarchical Takagi-Sugeno model designed using evolutionary algorithms. Performance comparison is done with neuro-fuzzy and genetic programming methods. Empirical results indicate that the techniques are robust and suitable for developing light weight risk assessment models, which could be integrated with intrusion detection and prevention systems.

Swarm Intelligence (SI) is an innovative distributed intelligent paradigm whereby the collective behaviors of unsophisticated individuals interacting locally with their environment cause coherent functional global patterns to emerge. In this paper, we model the scheduling problem for the multi-objective Flexible Job-shop Scheduling Problems (FJSP) and attempt to formulate and solve the problem using a Multi Particle Swarm Optimization (MPSO) approach. MPSO consists of multi-swarms of particles, which searches for the operation order update and machine selection. All the swarms search the optima synergistically and maintain the balance between diversity of particles and search space. We theoretically prove that the multi-swarm synergetic optimization algorithm converges with a probability of 1 towards the global optima. The details of the implementation for the multi-objective FJSP and the corresponding computational experiments are reported. The results indicate that the proposed algorithm is an efficient approach for the multi-objective FJSP, especially for large scale problems.

This paper proposes a rough set reduction scheme for Support Vector Machine (SVM). In the proposed scheme, SVM is used for the classification task based on the significance of each feature vector, while rough set is applied to improve feature selection and data reduction. Particle Swarm Optimization (PSO) is used to optimize the rough set feature reduction. The proposed approach is used to classify the brain cognitive state data sets from a cognitive Functional Magnetic Resonance Imaging (fMRI) experiment. Empirical results indicate that by using the proposed hybrid scheme it is feasible to achieve the desired classification very efficiently.

In this paper, we introduce turbulence in the particle swarm optimization (TPSO) and illustrate how this approach could be used for function optimization problems involving high dimensions. The proposed algorithm uses a minimum velocity threshold to control the velocity of particles. TPSO mechanism is similar to a turbulent pump, which supplies some power to the swarm system to explore new search spaces (better solutions). The minimum velocity threshold of the particles is tuned adoptively by using a fuzzy logic controller, which is further called as fuzzy adaptive TPSO (FATPSO). We evaluate and compare the performance of SPSO (Standard PSO), TPSO and FATPSO. Empirical results clearly demonstrate that the performance of SPSO degrades remarkably with the increase in the dimensions of the problem, while the influence is very little in the case of TPSO and FATPSO.

Crew grouping is an important problem and formulating a good solution always involves many challenges. For example, grouping soldiers intelligently to tank combat units, we should take into consideration the combined technical proficiency of the soldiers, the amount of military training, the units from which the soldiers come, their service age, personal background, etc. In this paper, we propose a hybrid fuzzy-genetic algorithm (FGA) approach to solve the crew grouping problem. Fuzzy logic based controllers are applied to fine-tune dynamically the crossover and mutation probability in the genetic algorithms, in an attempt to improve the algorithm performance. The FGA approach is compared with the standard genetic algorithm (SGA). Empirical results clearly demonstrates that while the SGA approach gives satisfactory solutions for the problem, the FGA method usually performs significantly better.