Concept prerequisite relation learning is used to identify dependency relations between knowledge concepts, which helps learners choose effective learning paths. Currently, most of the mainstream methods utilise deep learning algorithms to capture the prerequisite relations between concepts through supervised or semi-supervised learning. However, these methods are highly dependent on labelled data, which is scarce and costly to annotate in reality. To address this problem, we propose a framework called Weakly Supervised Enhanced Concept Prerequisite Relation Learning (WSECPRL). Specifically, we first generate an enhanced concept pseudo-relation graph without labeled data using the pre-trained language model and the large knowledge base as auxiliary information. Second, we propose an improved variational graph auto-encoder model to correctly determine the concept prerequisite relations. We incorporate a multi-head attention mechanism to enhance the representation learning capability of weakly supervised learning. The model reconstructs a directed graph into multiple undirected graphs by splitting the adjacency matrix and determines the direction of the concept prerequisite relation based on the strength of the dependency relation between concepts. Finally, experimental results on several publicly available datasets demonstrate the effectiveness of our proposed framework, with WSECPRL outperforming existing baseline models in terms of F1 scores and AUC.
{"title":"Enhancing Weak Supervision for Concept Prerequisite Relation Learning","authors":"Miao Zhang;Jiawei Wang;Kui Xiao;Zhifang Huang;Zhifei Li;Yan Zhang","doi":"10.1109/TBDATA.2025.3552330","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3552330","url":null,"abstract":"Concept prerequisite relation learning is used to identify dependency relations between knowledge concepts, which helps learners choose effective learning paths. Currently, most of the mainstream methods utilise deep learning algorithms to capture the prerequisite relations between concepts through supervised or semi-supervised learning. However, these methods are highly dependent on labelled data, which is scarce and costly to annotate in reality. To address this problem, we propose a framework called <underline>W</u>eakly <underline>S</u>upervised <underline>E</u>nhanced <underline>C</u>oncept <underline>P</u>rerequisite <underline>R</u>elation <underline>L</u>earning (WSECPRL). Specifically, we first generate an enhanced concept pseudo-relation graph without labeled data using the pre-trained language model and the large knowledge base as auxiliary information. Second, we propose an improved variational graph auto-encoder model to correctly determine the concept prerequisite relations. We incorporate a multi-head attention mechanism to enhance the representation learning capability of weakly supervised learning. The model reconstructs a directed graph into multiple undirected graphs by splitting the adjacency matrix and determines the direction of the concept prerequisite relation based on the strength of the dependency relation between concepts. Finally, experimental results on several publicly available datasets demonstrate the effectiveness of our proposed framework, with WSECPRL outperforming existing baseline models in terms of F1 scores and AUC.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 5","pages":"2643-2656"},"PeriodicalIF":5.7,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144934541","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-17DOI: 10.1109/TBDATA.2025.3552335
Lantian Xu;Rong-Hua Li;Dong Wen;Qiangqiang Dai;Guoren Wang
A signed graph is a graph where each edge receives a sign, positive or negative. The signed graph model has been used in many real applications, such as protein complex discovery and social network analysis. Finding cohesive subgraphs in signed graphs is a fundamental problem. A $k$-plex is a common model for cohesive subgraphs in which every vertex is adjacent to all but at most $k$ vertices within the subgraph. In this paper, we propose the model of size-constrained antagonistic $k$-plex in a signed graph. The proposed model guarantees that the resulting subgraph is a $k$-plex and can be divided into two sub-$k$-plexes, both of which have positive inner edges and negative outer edges. This paper aims to identify all maximal antagonistic $k$-plexes in a signed graph. Through rigorous analysis, we show that the problem is NP-Hardness. We propose a novel framework for maximal antagonistic $k$-plexes utilizing set enumeration. Efficiency is improved through pivot pruning and early termination based on the color bound. Preprocessing techniques based on degree and dichromatic graphs effectively narrow the search space before enumeration. Extensive experiments on real-world datasets demonstrate our algorithm’s efficiency, effectiveness, and scalability.
{"title":"Efficient Antagonistic $k$k-Plex Enumeration in Signed Graphs","authors":"Lantian Xu;Rong-Hua Li;Dong Wen;Qiangqiang Dai;Guoren Wang","doi":"10.1109/TBDATA.2025.3552335","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3552335","url":null,"abstract":"A signed graph is a graph where each edge receives a sign, positive or negative. The signed graph model has been used in many real applications, such as protein complex discovery and social network analysis. Finding cohesive subgraphs in signed graphs is a fundamental problem. A <inline-formula><tex-math>$k$</tex-math></inline-formula>-plex is a common model for cohesive subgraphs in which every vertex is adjacent to all but at most <inline-formula><tex-math>$k$</tex-math></inline-formula> vertices within the subgraph. In this paper, we propose the model of size-constrained antagonistic <inline-formula><tex-math>$k$</tex-math></inline-formula>-plex in a signed graph. The proposed model guarantees that the resulting subgraph is a <inline-formula><tex-math>$k$</tex-math></inline-formula>-plex and can be divided into two sub-<inline-formula><tex-math>$k$</tex-math></inline-formula>-plexes, both of which have positive inner edges and negative outer edges. This paper aims to identify all maximal antagonistic <inline-formula><tex-math>$k$</tex-math></inline-formula>-plexes in a signed graph. Through rigorous analysis, we show that the problem is NP-Hardness. We propose a novel framework for maximal antagonistic <inline-formula><tex-math>$k$</tex-math></inline-formula>-plexes utilizing set enumeration. Efficiency is improved through pivot pruning and early termination based on the color bound. Preprocessing techniques based on degree and dichromatic graphs effectively narrow the search space before enumeration. Extensive experiments on real-world datasets demonstrate our algorithm’s efficiency, effectiveness, and scalability.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 5","pages":"2587-2600"},"PeriodicalIF":5.7,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144998253","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-17DOI: 10.1109/TBDATA.2025.3552326
Hanpeng Liu;Shuoxi Zhang;Kun He
Knowledge distillation is widely used technique to transfer knowledge from a large pretrained teacher network to a small student network. However, training complex teacher models requires significant computational resources and storage. To address this, a growing area of research, known as self-knowledge distillation (Self-KD), aims to enhance the performance of a neural network by leveraging its own latent knowledge. Despite its potential, existing Self-KD methods often struggle to effectively extract and utilize the model's dark knowledge. In this work, we identify a consistency problem between feature layer and output layer, and propose a novel Self-KD approach called Residual Learning for Self-Knowledge Distillation (RSKD). Our method addresses this issue by enabling the last feature layer of the student model learn the residual gap between the outputs of the pseudo-teacher and the student. Additionally, we extend RSKD by allowing each intermediate feature layer of the student model to learn the residual gap between the corresponding deeper features of the pseudo-teacher and the student. Extensive experiments on various visual datasets demonstrate the effectiveness of the proposed method, which outperforms the state-of-the-art baselines.
{"title":"Residual Learning for Self-Knowledge Distillation: Enhancing Neural Networks Through Consistency Across Layers","authors":"Hanpeng Liu;Shuoxi Zhang;Kun He","doi":"10.1109/TBDATA.2025.3552326","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3552326","url":null,"abstract":"Knowledge distillation is widely used technique to transfer knowledge from a large pretrained teacher network to a small student network. However, training complex teacher models requires significant computational resources and storage. To address this, a growing area of research, known as self-knowledge distillation (Self-KD), aims to enhance the performance of a neural network by leveraging its own latent knowledge. Despite its potential, existing Self-KD methods often struggle to effectively extract and utilize the model's dark knowledge. In this work, we identify a consistency problem between feature layer and output layer, and propose a novel Self-KD approach called <bold>R</b>esidual Learning for <bold>S</b>elf-<bold>K</b>nowledge <bold>D</b>istillation (<bold>RSKD</b>). Our method addresses this issue by enabling the last feature layer of the student model learn the residual gap between the outputs of the pseudo-teacher and the student. Additionally, we extend RSKD by allowing each intermediate feature layer of the student model to learn the residual gap between the corresponding deeper features of the pseudo-teacher and the student. Extensive experiments on various visual datasets demonstrate the effectiveness of the proposed method, which outperforms the state-of-the-art baselines.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 5","pages":"2615-2627"},"PeriodicalIF":5.7,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144934466","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-17DOI: 10.1109/TBDATA.2025.3552343
Alvaro Cia-Mina;Jesus Lopez-Fidalgo;Weng Kee Wong
Huge data sets are widely available now and there is growing interest in selecting an optimal subsample from the full data set to improve inference efficiency and reduce labeling costs. We propose a new criterion called J–optimality, that builds upon a popular optimal selection criterion that minimizes the Random–X prediction error by additionally incorporating the joint distribution of the covariates. A key advantage of our approach is that we can relate the subsampling selection problem to that of finding an optimal approximate design under a convex criterion, where analytical tools for finding and studying them are already available. Consequently, the J–optimal subsampling method comes with theoretical results and theory-based algorithms for finding them. Simulation results and real data analysis show our proposed methods outperform current subsampling methods and the proposed algorithms can also adapt efficiently to select an optimal subsample from streaming data.
{"title":"Optimal Subdata Selection for Prediction Based on the Distribution of the Covariates","authors":"Alvaro Cia-Mina;Jesus Lopez-Fidalgo;Weng Kee Wong","doi":"10.1109/TBDATA.2025.3552343","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3552343","url":null,"abstract":"Huge data sets are widely available now and there is growing interest in selecting an optimal subsample from the full data set to improve inference efficiency and reduce labeling costs. We propose a new criterion called J–optimality, that builds upon a popular optimal selection criterion that minimizes the Random–X prediction error by additionally incorporating the joint distribution of the covariates. A key advantage of our approach is that we can relate the subsampling selection problem to that of finding an optimal approximate design under a convex criterion, where analytical tools for finding and studying them are already available. Consequently, the J–optimal subsampling method comes with theoretical results and theory-based algorithms for finding them. Simulation results and real data analysis show our proposed methods outperform current subsampling methods and the proposed algorithms can also adapt efficiently to select an optimal subsample from streaming data.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 5","pages":"2601-2614"},"PeriodicalIF":5.7,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10930599","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144934532","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-17DOI: 10.1109/TBDATA.2025.3552333
Lin Mu;Yide Cheng;Jun Shen;Yiwen Zhang;Hong Zhong
Event Extraction involves extracting event-related information such as event types and event arguments from context, which has long been tackled through well-designed neural networks or fine-tuned pre-trained language models. These approaches require substantial annotated data for tuning parameters and are resource-intensive. Recently, Prompting strategies with frozen parameters, such as Chain-of-Thought and Self-Consistency, have delivered success in NLP using LLMs by generating intermediate thought steps. However, they suffer from the challenge of error propagation and lack of interaction between different thoughts. In this paper, we propose Neural Network-based Prompting (NetPrompt), a novel network-structured prompting strategy for event extraction. The core idea behind NetPrompt is to imitate the excellent information integration capabilities of neural network structures. Specifically, we first decompose the event extraction problem into diverse intermediate subtasks, and each subtask is represented as a node in different layers of the network, the output of the nodes in the preceding layer is fed into the subsequent layer. Secondly, we propose pruning strategies to adapt the reasoning overhead to different problems. Finally, we have conducted extensive experiments on two widely used event extraction benchmarks to evaluate NetPrompt. The results demonstrated that NetPrompt significantly improved the event extraction performance compared to previous methods.
{"title":"NetPrompt: Neural Network Prompting Enhances Event Extraction in Large Language Models","authors":"Lin Mu;Yide Cheng;Jun Shen;Yiwen Zhang;Hong Zhong","doi":"10.1109/TBDATA.2025.3552333","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3552333","url":null,"abstract":"Event Extraction involves extracting event-related information such as event types and event arguments from context, which has long been tackled through well-designed neural networks or fine-tuned pre-trained language models. These approaches require substantial annotated data for tuning parameters and are resource-intensive. Recently, Prompting strategies with frozen parameters, such as Chain-of-Thought and Self-Consistency, have delivered success in NLP using LLMs by generating intermediate thought steps. However, they suffer from the challenge of error propagation and lack of interaction between different thoughts. In this paper, we propose <italic>Neural Network-based Prompting</i> (NetPrompt), a novel network-structured prompting strategy for event extraction. The core idea behind NetPrompt is to imitate the excellent information integration capabilities of neural network structures. Specifically, we first decompose the event extraction problem into diverse intermediate subtasks, and each subtask is represented as a node in different layers of the network, the output of the nodes in the preceding layer is fed into the subsequent layer. Secondly, we propose pruning strategies to adapt the reasoning overhead to different problems. Finally, we have conducted extensive experiments on two widely used event extraction benchmarks to evaluate NetPrompt. The results demonstrated that NetPrompt significantly improved the event extraction performance compared to previous methods.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 5","pages":"2628-2642"},"PeriodicalIF":5.7,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144934439","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The rapid advancement of location and information technologies has generated a significant volume of human mobility data, which has been extensively utilized in spatiotemporal recommendation systems, including personalized point-of-interest recommendation, route recommendation, and location-aware event recommendation. Achieving high-quality recommendation results necessitates excellent quality of input trajectory data. However, trajectories obtained from GPS-enabled devices often contain missing and erroneous data that is unevenly distributed over time and highly sparse, which significantly hampers the effectiveness spatiotemporal data analytics. Therefore, trajectory recovery plays an important role in spatiotemporal recommendation systems. The objective of trajectory recovery is to utilize historical trajectories to restore missing locations, providing high-quality data for spatiotemporal recommendation systems. The development of an effective trajectory recovery mechanism faces three major challenges: 1) Complex and multi-granularity transition patterns among different locations; 2) Difficulty in discovering spatio-temporal dependencies; and 3) Data sparsity and noise. To address these challenges, we propose an attentional model with spatio-temporal recurrent neural networks, ARMove, to recover human mobility from long and sparse trajectories. In ARMove, we first design a spatio-temporal weighted recurrent neural network to capture users’ long-term preferences. Next, we introduce a multi-granularity trajectory encoder to model complex transition patterns and multi-level periodicity of human mobility. An attention-based history aggregation module is proposed to leverage historical mobility information. Extensive evaluation results reveal that our model outperforms the state-of-the-art models, demonstrating its ability to reconstruct high-quality and fine-grained human mobility trajectories.
{"title":"Toward High-Quality Spatiotemporal Recommendation: Trajectory Recovery Based on Spatial and Temporal Dependencies","authors":"Yihao Zhao;Chenhao Wang;Hongyu Wang;Shunzhi Zhu;Lisi Chen","doi":"10.1109/TBDATA.2025.3570071","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3570071","url":null,"abstract":"The rapid advancement of location and information technologies has generated a significant volume of human mobility data, which has been extensively utilized in spatiotemporal recommendation systems, including personalized point-of-interest recommendation, route recommendation, and location-aware event recommendation. Achieving high-quality recommendation results necessitates excellent quality of input trajectory data. However, trajectories obtained from GPS-enabled devices often contain missing and erroneous data that is unevenly distributed over time and highly sparse, which significantly hampers the effectiveness spatiotemporal data analytics. Therefore, trajectory recovery plays an important role in spatiotemporal recommendation systems. The objective of trajectory recovery is to utilize historical trajectories to restore missing locations, providing high-quality data for spatiotemporal recommendation systems. The development of an effective trajectory recovery mechanism faces three major challenges: 1) Complex and multi-granularity transition patterns among different locations; 2) Difficulty in discovering spatio-temporal dependencies; and 3) Data sparsity and noise. To address these challenges, we propose an attentional model with spatio-temporal recurrent neural networks, ARMove, to recover human mobility from long and sparse trajectories. In ARMove, we first design a spatio-temporal weighted recurrent neural network to capture users’ long-term preferences. Next, we introduce a multi-granularity trajectory encoder to model complex transition patterns and multi-level periodicity of human mobility. An attention-based history aggregation module is proposed to leverage historical mobility information. Extensive evaluation results reveal that our model outperforms the state-of-the-art models, demonstrating its ability to reconstruct high-quality and fine-grained human mobility trajectories.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 4","pages":"1628-1639"},"PeriodicalIF":7.5,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144597721","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-14DOI: 10.1109/TBDATA.2025.3537217
Guang Yang;Jing Zhang;Giorgos Papanastasiou;Ge Wang;Dacheng Tao
{"title":"Editorial Emerging Horizons: The Rise of Large Language Models and Cross-Modal Generative AI","authors":"Guang Yang;Jing Zhang;Giorgos Papanastasiou;Ge Wang;Dacheng Tao","doi":"10.1109/TBDATA.2025.3537217","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3537217","url":null,"abstract":"","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 3","pages":"896-897"},"PeriodicalIF":7.5,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11003991","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143949196","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-20DOI: 10.1109/TBDATA.2025.3544126
Yi Xia;Gang Zhou;Junyong Luo;Mingjing Lan;Ningbo Huang
Reasoning over knowledge graphs has attracted considerable attention from researchers and is being widely applied to contribute question answering systems, recommender systems, and other information retrieval systems. However, existing reasoning methods tend to suffer from poor interpretability which is not consistent with human commonsense. The trustworthiness and reliability of the knowledge discover outcomes thus decreased as a result. Inspired by the process of human decision-making, we propose a commonsense-aware hierarchical framework called HDLH, which incorporates commonsense knowledge into hierarchical knowledge graph reasoning process with deep reinforcement learning. HDLH implements hierarchical reasoning process through exploration and exploitation sequentially by applying multi-agent reinforcement learning. Multiple agents in HDLH simulate the multi-level decision-making ability of humans, and reason hierarchically and reasonably to maintain its efficiency and interpretability. Moreover, commonsense knowledge is incorporated by means of the reward-shaping function, ultimately guiding the agent to reason more consistently with human perceptions and reduce the huge search space. We evaluated HDLH with various tasks on five real-world datasets. The experimental results reveal that HDLH achieves better performance compared with state-of-the-art baseline models.
{"title":"How to Decide Like Human? A Commonsense-Aware Hierarchical Framework for Knowledge Graph Reasoning","authors":"Yi Xia;Gang Zhou;Junyong Luo;Mingjing Lan;Ningbo Huang","doi":"10.1109/TBDATA.2025.3544126","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3544126","url":null,"abstract":"Reasoning over knowledge graphs has attracted considerable attention from researchers and is being widely applied to contribute question answering systems, recommender systems, and other information retrieval systems. However, existing reasoning methods tend to suffer from poor interpretability which is not consistent with human commonsense. The trustworthiness and reliability of the knowledge discover outcomes thus decreased as a result. Inspired by the process of human decision-making, we propose a commonsense-aware hierarchical framework called <italic>HDLH</i>, which incorporates commonsense knowledge into hierarchical knowledge graph reasoning process with deep reinforcement learning. <italic>HDLH</i> implements hierarchical reasoning process through exploration and exploitation sequentially by applying multi-agent reinforcement learning. Multiple agents in <italic>HDLH</i> simulate the multi-level decision-making ability of humans, and reason hierarchically and reasonably to maintain its efficiency and interpretability. Moreover, commonsense knowledge is incorporated by means of the reward-shaping function, ultimately guiding the agent to reason more consistently with human perceptions and reduce the huge search space. We evaluated <italic>HDLH</i> with various tasks on five real-world datasets. The experimental results reveal that <italic>HDLH</i> achieves better performance compared with state-of-the-art baseline models.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 5","pages":"2545-2556"},"PeriodicalIF":5.7,"publicationDate":"2025-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144934362","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-20DOI: 10.1109/TBDATA.2025.3544131
Jie Hu;Taichuan Zheng;Lilan Peng;Fei Teng;Shengdong Du;Tianrui Li
Traffic flow forecasting task plays an essential role in intelligent transportation systems. Accurately capturing the intricate spatio-temporal dependencies in traffic network signals is the core of precise prediction. Recently, a paradigm that models spatio-temporal dependencies through graph neural networks and time series models has become one of the most promising methods to solve this problem. However, existing methods still have limitations due to ineffectively modeling dynamic spatial dependencies and high time and space complexity. To address these issues, we propose a simplifying and powerful general spatio-temporal traffic flow forecasting model called LightST. Specifically, LightST first embeds temporal covariates and spatial position information to enhance the spatio-temporal modeling capabilities. Then, stacked temporal linear layers are introduced to capture temporal dependencies efficiently. Finally,we propose a concise adaptive spatio-temporal embedding graph convolution method to extract implicit spatial dependencies over time via dynamic graph convolution with adaptive spatio-temporal embedding graph generation. Extensive experiment results on four public traffic flow datasets demonstrate the superiority of our LightST concerning computational efficiency and prediction performance.
{"title":"LightST: A Simplifying Spatio-Temporal Graph Neural Network for Traffic Flow Forecasting","authors":"Jie Hu;Taichuan Zheng;Lilan Peng;Fei Teng;Shengdong Du;Tianrui Li","doi":"10.1109/TBDATA.2025.3544131","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3544131","url":null,"abstract":"Traffic flow forecasting task plays an essential role in intelligent transportation systems. Accurately capturing the intricate spatio-temporal dependencies in traffic network signals is the core of precise prediction. Recently, a paradigm that models spatio-temporal dependencies through graph neural networks and time series models has become one of the most promising methods to solve this problem. However, existing methods still have limitations due to ineffectively modeling dynamic spatial dependencies and high time and space complexity. To address these issues, we propose a simplifying and powerful general spatio-temporal traffic flow forecasting model called LightST. Specifically, LightST first embeds temporal covariates and spatial position information to enhance the spatio-temporal modeling capabilities. Then, stacked temporal linear layers are introduced to capture temporal dependencies efficiently. Finally,we propose a concise adaptive spatio-temporal embedding graph convolution method to extract implicit spatial dependencies over time via dynamic graph convolution with adaptive spatio-temporal embedding graph generation. Extensive experiment results on four public traffic flow datasets demonstrate the superiority of our LightST concerning computational efficiency and prediction performance.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 5","pages":"2517-2528"},"PeriodicalIF":5.7,"publicationDate":"2025-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144934557","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-20DOI: 10.1109/TBDATA.2025.3544129
Jing Xiao;Yu-Cheng Zou;Xiao-Ke Xu
Recently higher-order community detection based on network motifs has received increasing attention, because motif-based communities reflect not only mesoscale structures but also functional characteristics of real-life networks. In this study, we propose a Modularity Optimization method for Motif-based Community Detection (MOMCD). In order to approximate the global optimum in modularity optimization, an improved nature-inspired metaheuristic algorithm is proposed as optimization strategy. In addition, by comprehensively utilizing motif-based (higher-order) and edge-based (lower-order) structural information, a neighbor community modification operation and a local search operation are also designed to improve the quality of individuals and promote the convergence of MOMCD. Experimental results show that MOMCD is promising and competitive in identifying motif-based communities from synthetic and real-life networks, which outperforms state-of-the-art approaches in terms of quality and accuracy, and deepens our understanding of network structural and functional characteristics.
{"title":"Higher-Order Community Detection by Motif-Based Modularity Optimization","authors":"Jing Xiao;Yu-Cheng Zou;Xiao-Ke Xu","doi":"10.1109/TBDATA.2025.3544129","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3544129","url":null,"abstract":"Recently higher-order community detection based on network motifs has received increasing attention, because motif-based communities reflect not only mesoscale structures but also functional characteristics of real-life networks. In this study, we propose a Modularity Optimization method for Motif-based Community Detection (MOMCD). In order to approximate the global optimum in modularity optimization, an improved nature-inspired metaheuristic algorithm is proposed as optimization strategy. In addition, by comprehensively utilizing motif-based (higher-order) and edge-based (lower-order) structural information, a neighbor community modification operation and a local search operation are also designed to improve the quality of individuals and promote the convergence of MOMCD. Experimental results show that MOMCD is promising and competitive in identifying motif-based communities from synthetic and real-life networks, which outperforms state-of-the-art approaches in terms of quality and accuracy, and deepens our understanding of network structural and functional characteristics.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 5","pages":"2529-2544"},"PeriodicalIF":5.7,"publicationDate":"2025-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144934356","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：