Article
Automation & Control Systems
Yangfan Li, Cen Chen, Mingxing Duan, Zeng Zeng, Kenli Li
Summary: The recent trend focuses on using heterogeneous graphs for facilitating the application of deep learning in the Internet of Things, but existing models struggle to accurately represent complex semantics and attributes. To address this challenge, attention-aware encoder-decoder graph neural network called HGAED has been developed to improve accuracy using attention-based separate-and-merge method and encoder-decoder architecture. Extensive experiments show superior performance of HGAED over state-of-the-art baselines in fusing heterogeneous structures and contents of nodes hierarchically.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Engineering, Aerospace
Yong Sun, Wenan Tan, Li Huang, Na Xie, Ling Ruan, Li Da Xu
Summary: In this study, we propose an innovative integrated collaborative filtering framework that incorporates location-aware quality into a graph embedding model for reliable IoT service discovery. A privacy protection mechanism is also introduced to safeguard sensitive location information. Experimental results demonstrate the significant efficacy and reliability of our framework in large-scale IoT service discovery scenarios.
IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS
(2022)
Article
Automation & Control Systems
Yongheng Xing, Liang Hu, Xiaolu Zhang, Gang Wu, Feng Wang
Summary: In this article, a NMF-based heterogeneous graph embedding method called IoT nonnegative matrix factorization (IoT-NMF) is proposed for trigger-action programming (TAP) in Internet of Things (IoT) devices/web services. By mapping triggers and actions to an IoT heterogeneous information network, three structures that preserve heterogeneous relations are extracted, and IoT-NMF is used to simultaneously factorize these structures to obtain low-dimensional representation vectors. The proposed approach outperforms the state-of-the-art methods, as demonstrated using an if this then that (IFTTT) dataset.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Computer Science, Artificial Intelligence
Gen Li, Jason J. Jung
Summary: This study presents a novel approach for anomaly detection in IoT time series data and achieves better performance compared to other models on industrial IoT datasets.
Article
Multidisciplinary Sciences
Yasir Ali, Habib Ullah Khan
Summary: The supply chain management of COVID-19 vaccine is a complex task, and IoT technology is a suitable solution. This study proposes a decision making model to select the right IoT platform for the logistics and transportation process of COVID-19 vaccine. The model is validated and tested through surveys and shows high accuracy and reliability.
SCIENTIFIC REPORTS
(2023)
Article
Computer Science, Information Systems
Yuan Cheng, Baojiang Cui, Chen Chen, Thar Baker, Tao Qi
Summary: This paper proposes a new approach that combines ambient information with binary code information to generate an ACFG for discovering vulnerabilities from binaries accurately. The transformed graph is analyzed using a deep neural network. The experiment shows that this method outperforms state-of-the-art methods in terms of accuracy, efficiency, and scalability, with an average accuracy of over 80% on real-world vulnerability datasets.
COMPUTER COMMUNICATIONS
(2023)
Article
Computer Science, Information Systems
Amadou Ba, Karol Lynch, Joern Ploennigs, Ben Schaper, Christopher Lohse, Fabio Lorenzi
Summary: Efficient training of deep learning models for IoT systems requires understanding domain knowledge. Heterogeneous graph neural networks (HGNNs) are a promising approach to incorporate domain knowledge and improve model performance. However, encoding domain knowledge into HGNNs for IoT systems is challenging and manual. To overcome this, we propose a framework to automatically derive HGNN features by parsing equations in publications, and validate our approach with IoT use cases. Our approach significantly outperforms other techniques.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Computer Science, Information Systems
Seonghee Kim, Yongyoon Suh, Hakyeon Lee
Summary: This study explores user-created automation applets to connect IoT devices and applications, and builds an IoT application network using data from the IFTTT platform. By analyzing the trigger-action relationships and clustering the embedded nodes, different IoT usage patterns are identified. Predicting the IoT application network and generating feasible service scenarios offer practical implications for enhancing user experiences and developing new services.
INFORMATION PROCESSING & MANAGEMENT
(2022)
Article
Computer Science, Information Systems
Junyu Lu, Laurence T. T. Yang, Bing Guo, Qiang Li, Hong Su, Gongliang Li
Summary: To promote IoT data and semantic interoperability, an edge-cloud framework is proposed, where the edge end handles customized data processing tasks and the cloud end deals with semantic information processing. An entity tree embedding algorithm is presented to convert IoT entities and attributes into embedding vectors in a tree-structured way, capturing both the semantic and structural information. Evaluation results show that the Whitening algorithm is the best method to compress embedding vectors and the proposed embedding algorithm achieves better clustering results compared with the original entity embeddings and the uncompressed averaging method.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Engineering, Industrial
Lei Ren, Yingjie Li, Xiaokang Wang, Jin Cui, Lin Zhang
Summary: The Industrial Internet of Things (IIoT) plays a critical role in the development of digital servitization in smart manufacturing. However, there is a knowledge gap between different manufacturing fields, hindering the integration and utilization of industrial big data. To address this challenge, a Framework of Manufacturing Knowledge Graph (FMKG) is proposed, along with an attention-based graph embedding model (ABGE), to extract and complement relationships in the knowledge graph for efficient integration and leverage of industrial knowledge.
INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH
(2023)
Article
Computer Science, Information Systems
Jing Yang, Laurence T. Yang, Hao Wang, Yuan Gao, Huazhong Liu, Xia Xie
Summary: This article proposes a novel multirelational GAT framework for knowledge reasoning over heterogeneous graphs. By employing tensor and tensor operations to simulate rich interactions between mixed triples, entities, and relationships, and utilizing the Tucker model to compress parameters, the proposed TGAT model outperforms competitors in link prediction task on real-world graphs, significantly improving hits@1 accuracy by up to 7.6%.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Computer Science, Information Systems
Junyu Lu, Laurence T. Yang, Bing Guo, Qiang Li, Hong Su, Gongliang Li, Jun Tang
Summary: This article presents a dataspaces model utilizing distributed approaches for sustainable solution of semantic interoperability in IoT. By using attention-based entity embedding and relation recognition, the semantic information of IoT entities can be effectively calculated and represented. Experimental results demonstrate the effectiveness of the proposed approaches in semantic similarity and relation recognition.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Multidisciplinary Sciences
Hiep Xuan Huynh, Linh Nhut Tran, Nghia Duong-Trung
Summary: This paper presents a smart greenhouse IoT solution for the cultivation of Brassica Juncea, a mustard variety commonly grown in Vietnam. The research team successfully designed, constructed, and tested a 30m2 smart greenhouse, integrated diverse IoT technologies, and developed a web interface and mobile application for convenient monitoring and control.
Article
Computer Science, Theory & Methods
Dehai Zhang, Xiaoqiang Xia, Yun Yang, Po Yang, Cheng Xie, Menglong Cui, Qing Liu
Summary: With the advancement of IoT, Natural Language Processing has become crucial in Healthcare applications. Word similarity measurement is fundamental in semantic analysis, and a new method combining knowledge-graph-based and word-embedding-based similarity measures shows significant improvements.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2021)
Article
Computer Science, Information Systems
Qi Luo, Dongxiao Yu, Yanwei Zheng, Hao Sheng, Xiuzhen Cheng
Summary: This study investigates the use of deep graph generative models in simulating large-scale Internet of Things (IoT) networks and presents a variable graph autoencoder called Core-GAE that considers the coreness of nodes during network generation. Core-GAE preserves both local proximity similarity and global structural features when learning the structural features of graphs.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Review
Computer Science, Information Systems
Feng Wang, Liang Hu, Jin Zhou, Kuo Zhao
INTERNATIONAL JOURNAL OF DISTRIBUTED SENSOR NETWORKS
(2015)
Review
Engineering, Electrical & Electronic
Feng Wang, Liang Hu, Jiejun Hu, Jin Zhou, Kuo Zhao
IETE TECHNICAL REVIEW
(2017)
Article
Computer Science, Information Systems
Feng Wang, Liang Hu, Rui Sun, Jiejun Hu, Kuo Zhao
INFORMATION SCIENCES
(2018)
Article
Computer Science, Artificial Intelligence
Liang Hu, Yongheng Xing, Yanlei Gong, Kuo Zhao, Feng Wang
Article
Computer Science, Information Systems
Dongming Sun, Xiaolu Zhang, Kim-Kwang Raymond Choo, Liang Hu, Feng Wang
Summary: Digital investigations play a crucial role in criminal investigations and civil litigations due to the increasing prevalence of online communications. This paper introduces a Natural Language Processing (NLP)-based digital investigation platform, demonstrating its superiority over other existing methods through empirical comparisons.
COMPUTERS & SECURITY
(2021)
Article
Automation & Control Systems
Yongheng Xing, Liang Hu, Xiaolu Zhang, Gang Wu, Feng Wang
Summary: In this article, a NMF-based heterogeneous graph embedding method called IoT nonnegative matrix factorization (IoT-NMF) is proposed for trigger-action programming (TAP) in Internet of Things (IoT) devices/web services. By mapping triggers and actions to an IoT heterogeneous information network, three structures that preserve heterogeneous relations are extracted, and IoT-NMF is used to simultaneously factorize these structures to obtain low-dimensional representation vectors. The proposed approach outperforms the state-of-the-art methods, as demonstrated using an if this then that (IFTTT) dataset.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Computer Science, Information Systems
Xuyang Chen, Xiaolu Zhang, Michael Elliot, Xiaoyin Wang, Feng Wang
Summary: This paper summarizes the existing literature on the security issues of Trigger-Action Programming (TAP) in IoT smart home platforms, including logical errors in TAP rules and vulnerabilities in well-known TAP platforms. It also provides detection techniques and solutions based on different approaches, such as Model Checking and Natural Language Processing. Additionally, datasets from literature or publicly available sources are summarized for potential future TAP security research.
COMPUTERS & SECURITY
(2022)
Article
Computer Science, Information Systems
Feng Wang, Liang Hu, Dongdai Zhou, Rui Sun, Jiejun Hu, Kuo Zhao