Article
Computer Science, Artificial Intelligence
Lijun Dong, Hong Yao, Dan Li, Yi Wang, Shengwen Li, Qingzhong Liang
Summary: Graph embedding technique in artificial intelligence is important for processing complex graph data efficiently. Existing GNN models often have limitations in considering global topology information, leading to difficulties in distinguishing nodes with similar local topologies. The proposed AS-GNN model aims to address this issue by capturing global topology information based on the characteristics of complex networks.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Yaomin Chang, Chuan Chen, Weibo Hu, Zibin Zheng, Xiaocong Zhou, Shouzhi Chen
Summary: This paper introduces a novel Meta-path Extracted heterogeneous Graph Neural Network (MEGNN) that can effectively extract meaningful meta-paths in heterogeneous graphs, optimizing the representation learning of heterogeneous graphs and providing the model with interpretability and reliability.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Ruiwen Yuan, Yajing Wu, Yongqiang Tang, Junping Wang, Wensheng Zhang
Summary: Network representation learning has been successful in homogeneous network data analysis, but cannot be directly applied in multiplex networks. To address this, this paper proposes a novel model called Meta-path Infomax joint Structure Enhancement (MISE) for multiplex network representations. By utilizing the concept of meta-path and developing a meta-path infomax mechanism, the proposed model enhances the complementary information between different types of meta-paths and captures the implicit correlations between nodes to construct the latent graph structure. Experimental results show that MISE achieves a promising boost in performance on a variety of real-world datasets.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Yang Xiao, Pei Quan, MingLong Lei, Lingfeng Niu
Summary: This article proposes a method for representing heterogeneous graphs based on latent direct neighbors, which enhances neighborhood relationships, constructs semantic meta-paths, and generates more accurate predictions through the use of random walks and HodgeRank.
Review
Engineering, Multidisciplinary
Sufen Zhao, Rong Peng, Po Hu, Liansheng Tan
Summary: This article provides a comprehensive survey of key advancements in heterogeneous network embedding (HNE). It defines an encoder-decoder-based HNE model taxonomy and systematically reviews, compares, and summarizes various state-of-the-art HNE models. The article also analyzes the advantages and disadvantages of different model categories, identifies potentially competitive HNE frameworks, and discusses open issues and future directions.
CMES-COMPUTER MODELING IN ENGINEERING & SCIENCES
(2023)
Article
Computer Science, Artificial Intelligence
Xiaoyu Xu, Xiaoyu Shi, Mingsheng Shang
Summary: This research introduces a novel approach for effectively learning node representations in disassortative graph structures. The proposed method synthesizes the feature semantic space and the structure semantic space to find friendly neighbor spaces and learns the interrelationship between aggregated information and separated information using contrastive learning. Experimental results show that the proposed approach outperforms eight state-of-the-art GNN models in various graph mining tasks, demonstrating its superior graph representation ability.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Mathematics
Yuxuan Yang, Beibei Han, Zanxi Ran, Min Gao, Yingmei Wei
Summary: Graph-embedding learning aims to represent nodes in a graph network as low-dimensional dense vectors for practical analysis tasks. Graph neural networks based on deep learning have gained attention in this field, but they often have limitations in utilizing higher-order neighborhood information effectively and considering structural properties. To address these issues, we propose centrality encoding, attention mechanism, and random walk regularization to improve the node representation. Experimental results on benchmark datasets demonstrate that our model outperforms baseline methods in node-clustering and link prediction tasks, showing highly expressive graph embedding.
Article
Mathematics
Bin Wang, Yu Chen, Jinfang Sheng, Zhengkun He
Summary: Graph embedding is significant for graph analysis and research, and the emergence of graph neural networks has greatly improved the accuracy of graph embedding. However, existing methods neglect the influence of clusters, thus this paper proposes a new approach to incorporate cluster influence into graph embedding.
Article
Computer Science, Theory & Methods
Zhongming Han, Xuelian Jin, Haozhen Xing, Weijie Yang, Haitao Xiong
Summary: Given the heterogeneity of real-world networks and the low efficiency of directly mining networks, learning low-dimensional embeddings of nodes in heterogeneous information networks (HINs) becomes crucial. In this paper, we propose a framework called HGSAGE for learning similarity-preserved embeddings of nodes in HINs. HGSAGE addresses the problems of omitting global information and considering only first-order neighbors by incorporating mechanisms to capture global information, sample and aggregate features from immediate and mediate neighbors, and combine embeddings from different meta-paths. Experimental results show that HGSAGE outperforms baseline methods on multiple tasks in real-world heterogeneous networks. Moreover, HGSAGE has important application values in this research field.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2023)
Article
Computer Science, Artificial Intelligence
Lei Xu, Zhen-Yu He, Kai Wang, Chang-Dong Wang, Shu-Qiang Huang
Summary: In this paper, a novel semi-supervised GNN model called EMP is proposed, which performs explicit message-passing along meta-paths to accurately capture the semantic information of heterogeneous graphs. The model also introduces a split method for meta-paths and considers the mutual effect between various meta-paths in advance to accurately capture the semantic information of the whole set of meta-paths. Extensive experiments demonstrate the superiority of the proposed model.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Yayang Li, Shuqing Liang, Yuncheng Jiang
Summary: In this paper, the authors propose the Path Reliability-based Graph Attention Networks (PRGATs), a novel method that incorporates multi-hop neighboring context into attention score computation in Graph Neural Networks. Experimental results show that PRGATs outperforms existing methods in tasks such as node classification and graph universal adversarial attack.
Article
Mathematics
Xinglong Chang, Jianrong Wang, Rui Guo, Yingkui Wang, Weihao Li
Summary: This paper proposes a simple method to solve the problem of collapsing solutions in unsupervised graph representation learning. By using contrastive learning and an asymmetric design, the method achieves effective graph representation learning.
Article
Mathematics
Wenchuan Zhang, Weihua Ou, Weian Li, Jianping Gou, Wenjun Xiao, Bin Liu
Summary: Graph neural networks (GNNs) have gained attention for effectively processing graph-related data. Existing methods assume noise-free input graphs, which is frequently violated in real-world scenarios. To address this issue, we introduce virtual nodes and utilize Gumbel-Softmax to reweight edges, achieving differentiable graph structure learning (abbreviated as VN-GSL). Thorough evaluations on benchmark datasets demonstrate the superiority of our approach in terms of performance and efficiency. Our implementation will be publicly available.
Article
Computer Science, Artificial Intelligence
Jianian Zhu, Weixin Zeng, Junfeng Zhang, Jiuyang Tang, Xiang Zhao
Summary: Graph contrastive learning (GCL) offers a new perspective to reduce the reliance on labeled data for graph representation learning. Existing GCL methods utilize graph augmentation strategies such as node dropping and edge masking, to create augmented views of the original graph for contrastive learning. However, these methods are limited in capturing sufficient information for contrastive learning. This work proposes the use of hypergraph to establish a new view for graph contrastive learning, enabling the capture of high-order information and improving the quality of graph representations through the contrast between the hypergraph view and the original graph view.
INFORMATION FUSION
(2023)
Article
Computer Science, Artificial Intelligence
Li Zhang, Heda Song, Nikolaos Aletras, Haiping Lu
Summary: Graph convolutional network (GCN) is an effective neural network model for graph representation learning. This paper proposes a new node-feature convolutional (NFC) layer to tackle the limitations of standard GCN. Experimental results show that NFC-GCN outperforms state-of-the-art methods in node classification.
PATTERN RECOGNITION
(2022)
