Pub Date : 2025-01-27DOI: 10.1109/TBDATA.2025.3533891
Ezekiel B. Ouedraogo;Ammar Hawbani;Xingfu Wang;Zhi Liu;Liang Zhao;Mohammed A. A. Al-qaness;Saeed Hamood Alsamhi
Digital Twins are virtual representations of physical assets and systems that rely on effective Data Management to integrate, process, and analyze diverse data sources. This article comprehensively examines Data Management challenges, architectures, techniques, and applications in the context of Digital Twins. It explores key issues such as data heterogeneity, quality assurance, scalability, security, and interoperability. The paper outlines architectural approaches like centralized, distributed, cloud-based, and blockchain solutions and Data Management techniques for modeling, integration, fusion, quality management, and visualization. Domain-specific considerations across manufacturing, smart cities, healthcare, and other sectors are discussed. Finally, open research challenges related to standards, real-time data processing, intelligent Data Management, and ethical aspects are highlighted. By synthesizing the state-of-the-art, this review serves as a valuable reference for developing robust Data Management strategies that enable Digital Twin deployments.
{"title":"Digital Twin Data Management: A Comprehensive Review","authors":"Ezekiel B. Ouedraogo;Ammar Hawbani;Xingfu Wang;Zhi Liu;Liang Zhao;Mohammed A. A. Al-qaness;Saeed Hamood Alsamhi","doi":"10.1109/TBDATA.2025.3533891","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3533891","url":null,"abstract":"Digital Twins are virtual representations of physical assets and systems that rely on effective Data Management to integrate, process, and analyze diverse data sources. This article comprehensively examines Data Management challenges, architectures, techniques, and applications in the context of Digital Twins. It explores key issues such as data heterogeneity, quality assurance, scalability, security, and interoperability. The paper outlines architectural approaches like centralized, distributed, cloud-based, and blockchain solutions and Data Management techniques for modeling, integration, fusion, quality management, and visualization. Domain-specific considerations across manufacturing, smart cities, healthcare, and other sectors are discussed. Finally, open research challenges related to standards, real-time data processing, intelligent Data Management, and ethical aspects are highlighted. By synthesizing the state-of-the-art, this review serves as a valuable reference for developing robust Data Management strategies that enable Digital Twin deployments.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 5","pages":"2224-2243"},"PeriodicalIF":5.7,"publicationDate":"2025-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144990260","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 : 2025-01-15DOI: 10.1109/TBDATA.2025.3526356
{"title":"2024 Reviewers List*","authors":"","doi":"10.1109/TBDATA.2025.3526356","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3526356","url":null,"abstract":"","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 1","pages":"310-313"},"PeriodicalIF":7.5,"publicationDate":"2025-01-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10843074","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142993640","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 : 2025-01-14DOI: 10.1109/TBDATA.2025.3528727
Jing Wang;Dehui Kong;Baocai Yin
Weakly supervised temporal action localization (WTAL) aims to precisely locate action instances in given videos by video-level classification supervision, which is partly related to action classification. Most existing localization works directly utilize feature encoders pre-trained for video classification tasks to extract video features, resulting in non-targeted features that lead to incomplete or over-complete action localization. Therefore, we propose Generalized Contrast Learning Network (GCLNet), in which two novel strategies are proposed to improve the pre-trained features. First, to address the issue of over-completeness, GCLNet introduces text information with good context independence and category separability to enrich the expression of video features, as well as proposes a novel generalized contrastive learning approach for similarity metrics, which facilitates pulling closer the features belonging to the same category while pushing farther apart those from different categories. Consequently, it enables more compact intra-class feature learning and ensures accurate action localization. Second, to tackle the problem of incomplete, we exploit the respective advantages of RGB and Flow features in scene appearance and temporal motion expression, designing a hybrid attention strategy in GCLNet to enhance each channel features mutually. This process greatly improves the features through establishing cross-channel consensus. Finally, we conduct extensive experiments on THUMOS14 and ActivityNet1.2, respectively, and the results show that our proposed GCLNet can produce more representative action localization features.
{"title":"GCLNet: Generalized Contrastive Learning for Weakly Supervised Temporal Action Localization","authors":"Jing Wang;Dehui Kong;Baocai Yin","doi":"10.1109/TBDATA.2025.3528727","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3528727","url":null,"abstract":"Weakly supervised temporal action localization (WTAL) aims to precisely locate action instances in given videos by video-level classification supervision, which is partly related to action classification. Most existing localization works directly utilize feature encoders pre-trained for video classification tasks to extract video features, resulting in non-targeted features that lead to incomplete or over-complete action localization. Therefore, we propose Generalized Contrast Learning Network (GCLNet), in which two novel strategies are proposed to improve the pre-trained features. First, to address the issue of over-completeness, GCLNet introduces text information with good context independence and category separability to enrich the expression of video features, as well as proposes a novel generalized contrastive learning approach for similarity metrics, which facilitates pulling closer the features belonging to the same category while pushing farther apart those from different categories. Consequently, it enables more compact intra-class feature learning and ensures accurate action localization. Second, to tackle the problem of incomplete, we exploit the respective advantages of RGB and Flow features in scene appearance and temporal motion expression, designing a hybrid attention strategy in GCLNet to enhance each channel features mutually. This process greatly improves the features through establishing cross-channel consensus. Finally, we conduct extensive experiments on THUMOS14 and ActivityNet1.2, respectively, and the results show that our proposed GCLNet can produce more representative action localization features.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 5","pages":"2365-2375"},"PeriodicalIF":5.7,"publicationDate":"2025-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144990027","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}
Knowledge combination prediction involves analyzing current knowledge elements and their relationships, then forecasting how these elements, drawn from various fields, can be creatively combined to form new, innovative solutions. This process is critical for countries and businesses to understand future technology trends and promote innovation in an era of rapid scientific and technological advancement. Existing methods often overlook the integration of knowledge combinations from multiple views, along with their inherent heterophily and the dual “many-to-one” property, where a single knowledge combination can include multiple elements, and a single element may belong to various combinations. To this end, we propose a novel framework named Multi-view Heterogeneous HyperGNN for Heterophilic Knowledge Combination Prediction (H3KCP). Specifically, H3KCP first constructs a hypergraph reflecting the dual “many-to-one” property of knowledge combinations, where each hyperedge may contain several nodes and each node can also belong to multiple hyperedges. Next, the framework employs a multi-view fusion approach to model knowledge combinations, considering heterophily and integrating insights from co-occurrence, co-citation, and hierarchical structure-based views. Furthermore, our analysis of H3KCP from a spectral graph perspective offers insights into its rationality. Finally, extensive experiments on real-world patent datasets and the Open Academic Graph dataset validate the effectiveness and efficiency of our approach, yielding significant insights into knowledge combinations.
{"title":"Multi-View Heterogeneous HyperGNN for Heterophilic Knowledge Combination Prediction","authors":"Huijie Liu;Shulan Ruan;Han Wu;Zhenya Huang;Defu Lian;Qi Liu;Enhong Chen","doi":"10.1109/TBDATA.2025.3527216","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3527216","url":null,"abstract":"Knowledge combination prediction involves analyzing current knowledge elements and their relationships, then forecasting how these elements, drawn from various fields, can be creatively combined to form new, innovative solutions. This process is critical for countries and businesses to understand future technology trends and promote innovation in an era of rapid scientific and technological advancement. Existing methods often overlook the integration of knowledge combinations from multiple views, along with their inherent heterophily and the dual “many-to-one” property, where a single knowledge combination can include multiple elements, and a single element may belong to various combinations. To this end, we propose a novel framework named Multi-view <underline>H</u>eterogeneous <underline>H</u>yperGNN for <underline>H</u>eterophilic <underline>K</u>nowledge <underline>C</u>ombination <underline>P</u>rediction (H3KCP). Specifically, H3KCP first constructs a hypergraph reflecting the dual “many-to-one” property of knowledge combinations, where each hyperedge may contain several nodes and each node can also belong to multiple hyperedges. Next, the framework employs a multi-view fusion approach to model knowledge combinations, considering heterophily and integrating insights from co-occurrence, co-citation, and hierarchical structure-based views. Furthermore, our analysis of H3KCP from a spectral graph perspective offers insights into its rationality. Finally, extensive experiments on real-world patent datasets and the Open Academic Graph dataset validate the effectiveness and efficiency of our approach, yielding significant insights into knowledge combinations.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 5","pages":"2321-2337"},"PeriodicalIF":5.7,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144990029","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}
In the field of heterogeneous federated learning (FL), the key challenge is to efficiently and collaboratively train models across multiple clients with different data distributions, model structures, task objectives, computational capabilities, and communication resources. This diversity leads to significant heterogeneity, which increases the complexity of model training. In this paper, we first outline the basic concepts of heterogeneous FL and summarize the research challenges in FL in terms of five aspects: data, model, task, device and communication. In addition, we explore how existing state-of-the-art approaches cope with the heterogeneity of FL, and categorize and review these approaches at three different levels: data-level, model-level, and architecture-level. Subsequently, the paper extensively discusses privacy-preserving strategies in heterogeneous FL environments. Finally, the paper discusses current open issues and directions for future research, aiming to promote the further development of heterogeneous FL.
{"title":"Advances in Robust Federated Learning: A Survey With Heterogeneity Considerations","authors":"Chuan Chen;Tianchi Liao;Xiaojun Deng;Zihou Wu;Sheng Huang;Zibin Zheng","doi":"10.1109/TBDATA.2025.3527202","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3527202","url":null,"abstract":"In the field of heterogeneous federated learning (FL), the key challenge is to efficiently and collaboratively train models across multiple clients with different data distributions, model structures, task objectives, computational capabilities, and communication resources. This diversity leads to significant heterogeneity, which increases the complexity of model training. In this paper, we first outline the basic concepts of heterogeneous FL and summarize the research challenges in FL in terms of five aspects: data, model, task, device and communication. In addition, we explore how existing state-of-the-art approaches cope with the heterogeneity of FL, and categorize and review these approaches at three different levels: data-level, model-level, and architecture-level. Subsequently, the paper extensively discusses privacy-preserving strategies in heterogeneous FL environments. Finally, the paper discusses current open issues and directions for future research, aiming to promote the further development of heterogeneous FL.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 3","pages":"1548-1567"},"PeriodicalIF":7.5,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143949337","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 : 2025-01-08DOI: 10.1109/TBDATA.2025.3527230
Junwei Yin;Min Gao;Kai Shu;Zehua Zhao;Yinqiu Huang;Jia Wang
The wide dissemination of fake news has affected our lives in many aspects, making fake news detection important and attracting increasing attention. Existing approaches make substantial contributions in this field by modeling news from a single-modal or multi-modal perspective. However, these modal-based methods can result in sub-optimal outcomes as they ignore reader behaviors in news consumption and authenticity verification. For instance, they haven't taken into consideration the component-by-component reading process: from the headline, images, comments, to the body, which is essential for modeling news with more granularity. To this end, we propose an approach of Emulating the behaviors of readers (Ember) for fake news detection on social media, incorporating readers’ reading and verificating process to model news from the component perspective thoroughly. Specifically, we first construct intra-component feature extractors to emulate the behaviors of semantic analyzing on each component. Then, we design a module that comprises inter-component feature extractors and a sequence-based aggregator. This module mimics the process of verifying the correlation between components and the overall reading and verification sequence. Thus, Ember can handle the news with various components by emulating corresponding sequences. We conduct extensive experiments on nine real-world datasets, and the results demonstrate the superiority of Ember.
{"title":"Emulating Reader Behaviors for Fake News Detection","authors":"Junwei Yin;Min Gao;Kai Shu;Zehua Zhao;Yinqiu Huang;Jia Wang","doi":"10.1109/TBDATA.2025.3527230","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3527230","url":null,"abstract":"The wide dissemination of fake news has affected our lives in many aspects, making fake news detection important and attracting increasing attention. Existing approaches make substantial contributions in this field by modeling news from a single-modal or multi-modal perspective. However, these modal-based methods can result in sub-optimal outcomes as they ignore reader behaviors in news consumption and authenticity verification. For instance, they haven't taken into consideration the component-by-component reading process: from the headline, images, comments, to the body, which is essential for modeling news with more granularity. To this end, we propose an approach of <underline>Em</u>ulating the <underline>be</u>haviors of <underline>r</u>eaders (Ember) for fake news detection on social media, incorporating readers’ reading and verificating process to model news from the component perspective thoroughly. Specifically, we first construct intra-component feature extractors to emulate the behaviors of semantic analyzing on each component. Then, we design a module that comprises inter-component feature extractors and a sequence-based aggregator. This module mimics the process of verifying the correlation between components and the overall reading and verification sequence. Thus, Ember can handle the news with various components by emulating corresponding sequences. We conduct extensive experiments on nine real-world datasets, and the results demonstrate the superiority of Ember.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 5","pages":"2353-2364"},"PeriodicalIF":5.7,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144990028","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 : 2025-01-08DOI: 10.1109/TBDATA.2025.3527200
Zuodong Jin;Peng Qi;Muyan Yao;Dan Tao
With the widespread application of Big Data and intelligent information systems, the tenant has become the main form of most scenarios. As a data mining technique, the portrait has been widely used to provide targeted services. Therefore, we transfer the traditional user-driven portrait into tenant driven for churn prediction. To achieve it, this paper first proposes a three-layer architecture and defines the fine-grained features for creating portraits from the perspective of tenants. In a large-scale telecommunication industry dataset of 100,000 tenants, we construct the tenant portrait through the proposed framework, and analyze the influences of the defined features on churn possibility. Then, considering the information missing caused by privacy concerns, we come up with the CrossMatch, a portrait completion model based on semi-supervised and graph convolution, which combines the relation characteristics among tenants for recovering missing information. On this basis, we design the tenant churn prediction method based on a directed attention network. Moreover, we recover missing information on three public node datasets with CrossMatch, achieving around 1-2$%$ improvement. We then apply the directed attention network for churn prediction and achieve an Accuracy of 75.06$%$, Precision of 77.78$%$, and F1-score of 71.43$%$, which outperforms all the baselines.
{"title":"Portraying Fine-Grained Tenant Portrait for Churn Prediction Using Semi-Supervised Graph Convolution and Attention Network","authors":"Zuodong Jin;Peng Qi;Muyan Yao;Dan Tao","doi":"10.1109/TBDATA.2025.3527200","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3527200","url":null,"abstract":"With the widespread application of Big Data and intelligent information systems, the tenant has become the main form of most scenarios. As a data mining technique, the portrait has been widely used to provide targeted services. Therefore, we transfer the traditional user-driven portrait into tenant driven for churn prediction. To achieve it, this paper first proposes a three-layer architecture and defines the fine-grained features for creating portraits from the perspective of tenants. In a large-scale telecommunication industry dataset of 100,000 tenants, we construct the tenant portrait through the proposed framework, and analyze the influences of the defined features on churn possibility. Then, considering the information missing caused by privacy concerns, we come up with the <i>CrossMatch</i>, a portrait completion model based on semi-supervised and graph convolution, which combines the relation characteristics among tenants for recovering missing information. On this basis, we design the tenant churn prediction method based on a directed attention network. Moreover, we recover missing information on three public node datasets with <i>CrossMatch</i>, achieving around 1-2<inline-formula><tex-math>$%$</tex-math></inline-formula> improvement. We then apply the directed attention network for churn prediction and achieve an Accuracy of 75.06<inline-formula><tex-math>$%$</tex-math></inline-formula>, Precision of 77.78<inline-formula><tex-math>$%$</tex-math></inline-formula>, and F1-score of 71.43<inline-formula><tex-math>$%$</tex-math></inline-formula>, which outperforms all the baselines.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 5","pages":"2296-2307"},"PeriodicalIF":5.7,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144990167","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 : 2025-01-08DOI: 10.1109/TBDATA.2025.3527208
Hongyi Lin;Yang Liu;Liang Wang;Xiaobo Qu
The rapid development of Big Data and artificial intelligence (AI) is revolutionizing the automotive and transportation industries, leading to the creation of the Autonomous Modular Bus (AMB). Designed to address the key challenges of modern public transportation systems, the AMB adopts a modular dynamic assembly approach. However, existing research on the AMB predominantly focuses on operational aspects, whereas in-transit docking remains the primary obstacle to its commercial deployment. This challenge stems from the fact that current perception accuracy in autonomous vehicles is limited to the decimeter level, with insufficient capability to manage adverse weather and complex traffic conditions. To enable AMBs to achieve full-scenario autonomous driving capabilities, this paper reviews current perception technologies from three perspectives: single-vehicle single-sensor perception, multi-sensor fusion perception, and cooperative perception. It examines the characteristics of existing perception solutions and evaluates their applicability to AMB-specific requirements. Furthermore, considering the unique challenges of in-transit docking, this paper identifies and proposes four future research directions for advancing AMB perception systems as well as general autonomous driving technologies.
{"title":"Big Data-Driven Advancements and Future Directions in Vehicle Perception Technologies: From Autonomous Driving to Modular Buses","authors":"Hongyi Lin;Yang Liu;Liang Wang;Xiaobo Qu","doi":"10.1109/TBDATA.2025.3527208","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3527208","url":null,"abstract":"The rapid development of Big Data and artificial intelligence (AI) is revolutionizing the automotive and transportation industries, leading to the creation of the Autonomous Modular Bus (AMB). Designed to address the key challenges of modern public transportation systems, the AMB adopts a modular dynamic assembly approach. However, existing research on the AMB predominantly focuses on operational aspects, whereas in-transit docking remains the primary obstacle to its commercial deployment. This challenge stems from the fact that current perception accuracy in autonomous vehicles is limited to the decimeter level, with insufficient capability to manage adverse weather and complex traffic conditions. To enable AMBs to achieve full-scenario autonomous driving capabilities, this paper reviews current perception technologies from three perspectives: single-vehicle single-sensor perception, multi-sensor fusion perception, and cooperative perception. It examines the characteristics of existing perception solutions and evaluates their applicability to AMB-specific requirements. Furthermore, considering the unique challenges of in-transit docking, this paper identifies and proposes four future research directions for advancing AMB perception systems as well as general autonomous driving technologies.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 3","pages":"1568-1587"},"PeriodicalIF":7.5,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143949176","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 : 2025-01-08DOI: 10.1109/TBDATA.2024.3524830
Ren Li;Huazhong Liu;Xiaotong Zhou;Jiawei Wang;Jihong Ding;Laurence T. Yang;Hua Li;Yunfan Zhang
With the explosion of social media platforms, a substantial amount of data is generated from social information network. Tensor-based multi-modal clustering methods have been widely applied in various scenarios of social information network by mining potential correlative relationships from large-scale heterogeneous data. Nevertheless, the accuracy and efficiency of tensor-based multi-modal clustering methods are seriously restricted by noise data and the curse of dimensionality. Therefore, this paper presents a Tucker-based multi-modal clustering (TuMC) and an improved TuMC (ITuMC) to enhance the accuracy and efficiency of multi-modal clustering. First, we propose two Tucker-based attribute weight ranking learning approaches to calculate weight tensor efficiently. Then, we present a calculation approach for Tucker-based selective weighted tensor distance (SWTD) and a TuMC method. Meanwhile, an ITuMC method is explored by optimizing the calculation efficiency of the SWTD to further improve clustering speed. Finally, we present a Tucker-based multi-modal clustering and service framework for social information network. Extensive experimental results based on social Geolife GPS trajectory and electricity consumption datasets demonstrate that the TuMC and ITuMC methods can cluster multi-source heterogeneous data with both higher accuracy and efficiency under complex social information network by DVI, AR and execution time measurement.
{"title":"Tucker-Based High-Accuracy Multi-Modal Clustering for Social Information Network","authors":"Ren Li;Huazhong Liu;Xiaotong Zhou;Jiawei Wang;Jihong Ding;Laurence T. Yang;Hua Li;Yunfan Zhang","doi":"10.1109/TBDATA.2024.3524830","DOIUrl":"https://doi.org/10.1109/TBDATA.2024.3524830","url":null,"abstract":"With the explosion of social media platforms, a substantial amount of data is generated from social information network. Tensor-based multi-modal clustering methods have been widely applied in various scenarios of social information network by mining potential correlative relationships from large-scale heterogeneous data. Nevertheless, the accuracy and efficiency of tensor-based multi-modal clustering methods are seriously restricted by noise data and the curse of dimensionality. Therefore, this paper presents a Tucker-based multi-modal clustering (TuMC) and an improved TuMC (ITuMC) to enhance the accuracy and efficiency of multi-modal clustering. First, we propose two Tucker-based attribute weight ranking learning approaches to calculate weight tensor efficiently. Then, we present a calculation approach for Tucker-based selective weighted tensor distance (SWTD) and a TuMC method. Meanwhile, an ITuMC method is explored by optimizing the calculation efficiency of the SWTD to further improve clustering speed. Finally, we present a Tucker-based multi-modal clustering and service framework for social information network. Extensive experimental results based on social Geolife GPS trajectory and electricity consumption datasets demonstrate that the TuMC and ITuMC methods can cluster multi-source heterogeneous data with both higher accuracy and efficiency under complex social information network by DVI, AR and execution time measurement.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 4","pages":"1677-1691"},"PeriodicalIF":7.5,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144597724","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 : 2025-01-08DOI: 10.1109/TBDATA.2025.3527238
Yao Liu;Yongfei Zhang
With the increasing amount of data in various domains, knowledge graphs (KGs) have become powerful tools for representing complex and heterogeneous information in a structured way, and for extracting valuable information from knowledge graphs through embedding techniques to support downstream tasks such as recommendation and Q&A systems. Knowledge graphs consist of triples that are continuously added as knowledge is updated. However, most existing embedding models are designed for static graphs, requiring the entire model to be retrained for each update, which is time-consuming. Existing global dynamic embedding models focus on exploiting the structural and relational information of the whole graph to achieve embedding quality, resulting in reduced dynamic efficiency. To address this problem, we propose a relational clustering-based parallel space model in which knowledge from different domains is embedded in different subspaces, allowing each subspace to focus on the data characteristics of a specific domain, thereby improving the quality of knowledge. Second, the new data only affects some subspaces but not the performance of other spaces, improving the model's adaptability to dynamics. Furthermore, we employ two incremental approaches based on the type of added data to improve the efficiency of dynamic embedding while ensuring that the added data preserves the characteristics of the parallel space. The experimental results show that the dynamic embedding efficiency of our model is improved by an average of 50.3% compared to the SOTA dynamic embedding model for the link prediction task. Particularly on FB15K, our model not only improves the efficiency by 41% but also increases the accuracy by 7.5%, demonstrating the accuracy and efficiency of our model.
{"title":"Relational Clustering-Based Parallel Spaces Construction and Embedding for Dynamic Knowledge Graph","authors":"Yao Liu;Yongfei Zhang","doi":"10.1109/TBDATA.2025.3527238","DOIUrl":"https://doi.org/10.1109/TBDATA.2025.3527238","url":null,"abstract":"With the increasing amount of data in various domains, knowledge graphs (KGs) have become powerful tools for representing complex and heterogeneous information in a structured way, and for extracting valuable information from knowledge graphs through embedding techniques to support downstream tasks such as recommendation and Q&A systems. Knowledge graphs consist of triples that are continuously added as knowledge is updated. However, most existing embedding models are designed for static graphs, requiring the entire model to be retrained for each update, which is time-consuming. Existing global dynamic embedding models focus on exploiting the structural and relational information of the whole graph to achieve embedding quality, resulting in reduced dynamic efficiency. To address this problem, we propose a relational clustering-based parallel space model in which knowledge from different domains is embedded in different subspaces, allowing each subspace to focus on the data characteristics of a specific domain, thereby improving the quality of knowledge. Second, the new data only affects some subspaces but not the performance of other spaces, improving the model's adaptability to dynamics. Furthermore, we employ two incremental approaches based on the type of added data to improve the efficiency of dynamic embedding while ensuring that the added data preserves the characteristics of the parallel space. The experimental results show that the dynamic embedding efficiency of our model is improved by an average of 50.3% compared to the SOTA dynamic embedding model for the link prediction task. Particularly on FB15K, our model not only improves the efficiency by 41% but also increases the accuracy by 7.5%, demonstrating the accuracy and efficiency of our model.","PeriodicalId":13106,"journal":{"name":"IEEE Transactions on Big Data","volume":"11 5","pages":"2308-2320"},"PeriodicalIF":5.7,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144990170","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：