Chenxu Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Chenxu Wang has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Xin Wang
Southwest Petroleum University
4 shared publications - Zhao Li
College of Intelligence and Computing, Tianjin University, Tianjin, China
4 shared publications - Zirui Chen
College of Intelligence and Computing, Tianjin University, Tianjin, China
4 shared publications - Chi Zhang
Alibaba Group, Hangzhou, China
2 shared publications - Jianxin Li
Edith Cowan University
2 shared publications - Jianzhi Shao
Alibaba Group, Hangzhou, China
2 shared publications
Latest Publications
- Open Access
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 4993-4998
TRACE: Term-level Reasoning And Chain-of-thought Enhanced distillation for E-commerce Multi-modal Relevance Learning
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 4993-4998
Open AccessTRACE: Term-level Reasoning And Chain-of-thought Enhanced distillation for E-commerce Multi-modal Relevance Learning
- Open Access
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 4965-4970
STAR: Staged Training with Aligned Reinforcement Learning and Multi-Faceted Distillation for Interpretable E-commerce Relevance
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 4965-4970
Open AccessSTAR: Staged Training with Aligned Reinforcement Learning and Multi-Faceted Distillation for Interpretable E-commerce Relevance
- 13 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 36 (8) , 3879-3892
HJE: Joint Convolutional Representation Learning for Knowledge Hypergraph Completion
Knowledge hypergraph representation learning , which projects entities and$n$ -ary relations into a low-dimensional vector space, remains a challenging area to be explored despite the ubiquity of$n$ -ary relational facts in the real world. Current methods are always extensions of those used for knowledge graphs with shallow or deep structures. However, shallow and linear models limit the extraction capacity of the latent knowledge, while deep and non-linear models lead to the overabundance of parameters. In this paper, we propose a novel knowledge hypergraph completion model called HJE, which utilizes the powerful capability of convolutional neural networks for efficient representation learning. Interaction-enhanced 3D convolution and relation-aware 2D convolution are jointly utilized by HJE to extract explicit and implicit global knowledge and semantic information effectively without compromising the translation property of the model. Moreover, HJE constructs a unified learnable embedding matrix to capture entity position information in knowledge tuples. The entity mask mechanism can naturally couple the multilinear scoring approach for$n$ -ary facts to speed up the training convergence of the model. Extensive experimental results on real datasets of knowledge hypergraphs and knowledge graphs demonstrate the superior performance of HJE compared with state-of-the-art baselines.- 13 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 36 (8) , 3879-3892
HJE: Joint Convolutional Representation Learning for Knowledge Hypergraph Completion
Knowledge hypergraph representation learning , which projects entities and$n$ -ary relations into a low-dimensional vector space, remains a challenging area to be explored despite the ubiquity of$n$ -ary relational facts in the real world. Current methods are always extensions of those used for knowledge graphs with shallow or deep structures. However, shallow and linear models limit the extraction capacity of the latent knowledge, while deep and non-linear models lead to the overabundance of parameters. In this paper, we propose a novel knowledge hypergraph completion model called HJE, which utilizes the powerful capability of convolutional neural networks for efficient representation learning. Interaction-enhanced 3D convolution and relation-aware 2D convolution are jointly utilized by HJE to extract explicit and implicit global knowledge and semantic information effectively without compromising the translation property of the model. Moreover, HJE constructs a unified learnable embedding matrix to capture entity position information in knowledge tuples. The entity mask mechanism can naturally couple the multilinear scoring approach for$n$ -ary facts to speed up the training convergence of the model. Extensive experimental results on real datasets of knowledge hypergraphs and knowledge graphs demonstrate the superior performance of HJE compared with state-of-the-art baselines.- Open Access
- 30 April 2023
- Published by Association for Computing Machinery (ACM)
- p. 188-198
HyConvE: A Novel Embedding Model for Knowledge Hypergraph Link Prediction with Convolutional Neural Networks
Knowledge hypergraph embedding, which projects entities and n-ary relations into a low-dimensional continuous vector space to predict missing links, remains a challenging area to be explored despite the ubiquity of n-ary relational facts in the real world. Currently, knowledge hypergraph link prediction methods are essentially simple extensions of those used in knowledge graphs, where n-ary relational facts are decomposed into different subelements. Convolutional neural networks have been shown to have remarkable information extraction capabilities in previous work on knowledge graph link prediction. In this paper, we propose a novel embedding-based knowledge hypergraph link prediction model named HyConvE, which exploits the powerful learning ability of convolutional neural networks for effective link prediction. Specifically, we employ 3D convolution to capture the deep interactions of entities and relations to efficiently extract explicit and implicit knowledge in each n-ary relational fact without compromising its translation property. In addition, appropriate relation and position-aware filters are utilized sequentially to perform two-dimensional convolution operations to capture the intrinsic patterns and position information in each n-ary relation, respectively. Extensive experimental results on real datasets of knowledge hypergraphs and knowledge graphs demonstrate the superior performance of HyConvE compared with state-of-the-art baselines.
- 30 April 2023
- Published by Association for Computing Machinery (ACM)
- p. 188-198
Open AccessHyConvE: A Novel Embedding Model for Knowledge Hypergraph Link Prediction with Convolutional Neural Networks
Knowledge hypergraph embedding, which projects entities and n-ary relations into a low-dimensional continuous vector space to predict missing links, remains a challenging area to be explored despite the ubiquity of n-ary relational facts in the real world. Currently, knowledge hypergraph link prediction methods are essentially simple extensions of those used in knowledge graphs, where n-ary relational facts are decomposed into different subelements. Convolutional neural networks have been shown to have remarkable information extraction capabilities in previous work on knowledge graph link prediction. In this paper, we propose a novel embedding-based knowledge hypergraph link prediction model named HyConvE, which exploits the powerful learning ability of convolutional neural networks for effective link prediction. Specifically, we employ 3D convolution to capture the deep interactions of entities and relations to efficiently extract explicit and implicit knowledge in each n-ary relational fact without compromising its translation property. In addition, appropriate relation and position-aware filters are utilized sequentially to perform two-dimensional convolution operations to capture the intrinsic patterns and position information in each n-ary relation, respectively. Extensive experimental results on real datasets of knowledge hypergraphs and knowledge graphs demonstrate the superior performance of HyConvE compared with state-of-the-art baselines.
- Open Access
- 30 April 2023
- Published by Association for Computing Machinery (ACM)
- p. 115-118
EnhancE: Enhanced Entity and Relation Embedding for Knowledge Hypergraph Link Prediction
Knowledge Hypergraphs, as the generalization of knowledge graphs, have attracted increasingly widespread attention due to their friendly compatibility with real-world facts. However, link prediction in knowledge hypergraph is still an underexplored field despite the ubiquity of n-ary facts in the real world. Several recent representative embedding-based knowledge hypergraph link prediction methods have proven to be effective in a series of benchmarks, however, they only consider the position (or role) information, ignoring the neighborhood structure among entities and rich semantic information within each fact. To this end, we propose a model named EnhancE for effective link prediction in knowledge hypergraphs. On the one hand, a more expressive entity representation is obtained with both position and neighborhood information added to the initial embedding. On the other hand, rich semantic information of the involved entities within each tuple is incorporated into relation embedding for enhanced representation. Extensive experimental results over real datasets of both knowledge hypergraph and knowledge graph demonstrate the excellent performance of EnhancE compared with a variety of state-of-the-art baselines.
- 30 April 2023
- Published by Association for Computing Machinery (ACM)
- p. 115-118
Open AccessEnhancE: Enhanced Entity and Relation Embedding for Knowledge Hypergraph Link Prediction
Knowledge Hypergraphs, as the generalization of knowledge graphs, have attracted increasingly widespread attention due to their friendly compatibility with real-world facts. However, link prediction in knowledge hypergraph is still an underexplored field despite the ubiquity of n-ary facts in the real world. Several recent representative embedding-based knowledge hypergraph link prediction methods have proven to be effective in a series of benchmarks, however, they only consider the position (or role) information, ignoring the neighborhood structure among entities and rich semantic information within each fact. To this end, we propose a model named EnhancE for effective link prediction in knowledge hypergraphs. On the one hand, a more expressive entity representation is obtained with both position and neighborhood information added to the initial embedding. On the other hand, rich semantic information of the involved entities within each tuple is incorporated into relation embedding for enhanced representation. Extensive experimental results over real datasets of both knowledge hypergraph and knowledge graph demonstrate the excellent performance of EnhancE compared with a variety of state-of-the-art baselines.
- 10 February 2023
- Published by Springer Nature
- p. 323-337
POSE: A Positional Embedding Model for Knowledge Hypergraph Link Prediction
- 10 February 2023
- Published by Springer Nature
- p. 323-337
POSE: A Positional Embedding Model for Knowledge Hypergraph Link Prediction