Wei Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Wei Wang has a total of 36 co-authors. Here are the top co-authors by number of shared publications:
- Hai-Tao Zheng
Tsinghua University
2 shared publications - Arun Kumar Sangaiah
National Yunlin University of Science and Technology
1 shared publication - Bin Hu
Key Laboratory of Brain Health Intelligent Evaluation and Intervention, Ministry of Education, Beijing Institute of Technology, Beijing, China
1 shared publication - Canrong Zhang
Logistics Engineering and Simulation Laboratory, Graduate School at Shenzhen, Tsinghua University, Shenzhen 518055, China
1 shared publication Changlong Yu
Key Laboratory of Brain Health Intelligent Evaluation and Intervention of Ministry of Education, School of Medical Technology, Bejing Institute of Technology (BIT), China
1 shared publication- Changlong Yu
Amazon, Palo Alto, USA
1 shared publication
Latest Publications
- Open Access
- 22 October 2026
- Published by Elsevier in SSRN Electronic Journal
BIM-AuditBench: A Granularity-Controlled Benchmark for Evaluating Large Language Models in Building Information Modeling (BIM)-based Code Compliance Checking
Large language models (LLMs) are increasingly considered as reasoning components in intelligent Building Information Modeling (BIM)-based compliance systems. Reliable deployment requires models to retrieve project-specific Industry Foundation Classes (IFC) evidence, operate external tools, integrate regulatory and engineering information, and support valid compliance decisions. Existing evaluations often examine domain knowledge, regulatory interpretation, BIM retrieval, code generation, or tool use separately, limiting diagnosis of the complete evidence-grounded reasoning process. This paper presents BIM-AuditBench, a granularity-controlled benchmark for evaluating LLMs in BIM-based code compliance checking. It contains 2,858 quality-controlled question–answer (QA) items grounded in three IFC model settings, 25 building-industry standards, enterprise knowledge, and auxiliary engineering data. A capability–granularity taxonomy spans three capability tracks and three reasoning levels. A 54-tool IfcOpenShell-based plan–call–answer protocol provides structured BIM evidence while separating evidence acquisition from final compliance judgment. The QA-item quality evaluation and model-response scoring modules are separately validated against domain-expert scores. Evaluation of thirteen LLMs on 590 held-out questions shows a clear gap between knowledge-oriented performance and BIM-grounded or engineering-decision reasoning. Tool-use and error analyses diagnose reliability bottlenecks in evidence retrieval, tool control, regulatory interpretation, structured-data reasoning, and decision synthesis.
- 22 October 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessBIM-AuditBench: A Granularity-Controlled Benchmark for Evaluating Large Language Models in Building Information Modeling (BIM)-based Code Compliance Checking
Large language models (LLMs) are increasingly considered as reasoning components in intelligent Building Information Modeling (BIM)-based compliance systems. Reliable deployment requires models to retrieve project-specific Industry Foundation Classes (IFC) evidence, operate external tools, integrate regulatory and engineering information, and support valid compliance decisions. Existing evaluations often examine domain knowledge, regulatory interpretation, BIM retrieval, code generation, or tool use separately, limiting diagnosis of the complete evidence-grounded reasoning process. This paper presents BIM-AuditBench, a granularity-controlled benchmark for evaluating LLMs in BIM-based code compliance checking. It contains 2,858 quality-controlled question–answer (QA) items grounded in three IFC model settings, 25 building-industry standards, enterprise knowledge, and auxiliary engineering data. A capability–granularity taxonomy spans three capability tracks and three reasoning levels. A 54-tool IfcOpenShell-based plan–call–answer protocol provides structured BIM evidence while separating evidence acquisition from final compliance judgment. The QA-item quality evaluation and model-response scoring modules are separately validated against domain-expert scores. Evaluation of thirteen LLMs on 590 held-out questions shows a clear gap between knowledge-oriented performance and BIM-grounded or engineering-decision reasoning. Tool-use and error analyses diagnose reliability bottlenecks in evidence retrieval, tool control, regulatory interpretation, structured-data reasoning, and decision synthesis.
- Open Access
- 1 January 2026
- Published by KEAI Publishing in AI Open
- Vol. 7, 276-288
DCF: Uniformity optimization for dual collaborative filtering
- 1 January 2026
- Published by KEAI Publishing in AI Open
- Vol. 7, 276-288
Open AccessDCF: Uniformity optimization for dual collaborative filtering
- 21 May 2025
- Published by Elsevier in Expert Systems with Applications
- Vol. 288
Event AutoAugment: Word-level data augmentation for textual event detection
- 21 May 2025
- Published by Elsevier in Expert Systems with Applications
- Vol. 288
Event AutoAugment: Word-level data augmentation for textual event detection
- 26 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. PP (10414347) , 1-14
Decider: A Dual-System Rule-Controllable Decoding Framework for Language Generation
Constrained decoding approaches aim to control the meaning or style of text generated by a Pre-trained Language Model (PLM) using specific target words during inference. However, these methods often guide plausible continuations by greedily selecting targets, which, while completing the task, may disrupt the natural patterns of human language generation. In this work, we propose a novel decoding framework, DECIDER, which enables us to program rules on how we complete tasks to control a PLM. Differing from previous work, our framework transforms the encouragement of target words into the encouragement of all words that satisfy the rule. Specifically, DECIDER is a dual system where a PLM is equipped with a First-Order- Logic (FOL) reasoner to express and evaluate the rules, and a decision function to merge the outputs from both systems to steer the generation. Experiments on CommonGen and PersonaChat demonstrate that DECIDER can effectively follow given rules to achieve generation tasks in a more human-like manner.
- 26 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. PP (10414347) , 1-14
Decider: A Dual-System Rule-Controllable Decoding Framework for Language Generation
Constrained decoding approaches aim to control the meaning or style of text generated by a Pre-trained Language Model (PLM) using specific target words during inference. However, these methods often guide plausible continuations by greedily selecting targets, which, while completing the task, may disrupt the natural patterns of human language generation. In this work, we propose a novel decoding framework, DECIDER, which enables us to program rules on how we complete tasks to control a PLM. Differing from previous work, our framework transforms the encouragement of target words into the encouragement of all words that satisfy the rule. Specifically, DECIDER is a dual system where a PLM is equipped with a First-Order- Logic (FOL) reasoner to express and evaluate the rules, and a decision function to merge the outputs from both systems to steer the generation. Experiments on CommonGen and PersonaChat demonstrate that DECIDER can effectively follow given rules to achieve generation tasks in a more human-like manner.
- 15 September 2022
- Published by Springer Nature
- p. 593-604
Context Reasoning Attention Network: Generating Plausible Distractors for Multi-choice Questions
- 15 September 2022
- Published by Springer Nature
- p. 593-604
Context Reasoning Attention Network: Generating Plausible Distractors for Multi-choice Questions
- Open Access
Consistency and Coherency Enhanced Story Generation
Story generation is a challenging task, which demands to maintain consistency of the plots and characters throughout the story. Previous works have shown that GPT2, a large-scale language model, has achieved good performance on story generation. However, we observe that several serious issues still exist in the stories generated by GPT2 which can be categorized into two folds: consistency and coherency. In terms of consistency, on one hand, GPT2 cannot guarantee the consistency of the plots explicitly. On the other hand, the generated stories usually contain coreference errors. In terms of coherency, GPT2 does not take account of the discourse relations between sentences of stories directly. To enhance the consistency and coherency of the generated stories, we propose a two-stage generation framework, where the first stage is to organize the story outline which depicts the story plots and events, and the second stage is to expand the outline into a complete story. Therefore the plots consistency can be controlled and guaranteed explicitly. In addition, coreference supervision signals are incorporated to reduce coreference errors and improve the coreference consistency. Moreover, we design an auxiliary task of discourse relation modeling to improve the coherency of the generated stories. Experimental results on a story dataset show that our model outperforms the baseline approaches in terms of both automatic metrics and human evaluation.
Open AccessConsistency and Coherency Enhanced Story Generation
Story generation is a challenging task, which demands to maintain consistency of the plots and characters throughout the story. Previous works have shown that GPT2, a large-scale language model, has achieved good performance on story generation. However, we observe that several serious issues still exist in the stories generated by GPT2 which can be categorized into two folds: consistency and coherency. In terms of consistency, on one hand, GPT2 cannot guarantee the consistency of the plots explicitly. On the other hand, the generated stories usually contain coreference errors. In terms of coherency, GPT2 does not take account of the discourse relations between sentences of stories directly. To enhance the consistency and coherency of the generated stories, we propose a two-stage generation framework, where the first stage is to organize the story outline which depicts the story plots and events, and the second stage is to expand the outline into a complete story. Therefore the plots consistency can be controlled and guaranteed explicitly. In addition, coreference supervision signals are incorporated to reduce coreference errors and improve the coreference consistency. Moreover, we design an auxiliary task of discourse relation modeling to improve the coherency of the generated stories. Experimental results on a story dataset show that our model outperforms the baseline approaches in terms of both automatic metrics and human evaluation.
- 28 November 2019
- Published by Springer Nature
- p. 714-723
A Universal Fusion Strategy for Image Super-Resolution Jointly from External and Internal Examples
- 28 November 2019
- Published by Springer Nature
- p. 714-723
A Universal Fusion Strategy for Image Super-Resolution Jointly from External and Internal Examples
- Open Access
- 17 November 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 6, 702-710
Automatic Generation of News Comments Based on Gated Attention Neural Networks
With the development of recurrent neural networks (RNN), various natural language generation (NLG) tasks have boomed in the past few years, such as response generation in conversation and poetry generation. However, automatic generation of news comments is anew, challenging and not well-studied task in NLG. Different from other NLG tasks, this task requires the contextual relevance between comments and news. In addition, we need to generate diversified comments, because different people usually have different opinions on the same news in the real world. In this paper, we propose a gated attention neural network model (GANN) to generate news comments. To address the problem of contextual relevance, we introduce the gated attention mechanism to use news context self-adaptively and selectively. To ensure the diversity of comments, we use random sample and relevance control to generate comments with different topics and degrees of relevance. Moreover, we apply generative adversarial nets to improve GANN. Automatic evaluation with perplexity score reveals that GANN outperforms the existing comment generation methods. Human evaluation proves that the generated news comments are close to human comments.
- 17 November 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 6, 702-710
Open AccessAutomatic Generation of News Comments Based on Gated Attention Neural Networks
With the development of recurrent neural networks (RNN), various natural language generation (NLG) tasks have boomed in the past few years, such as response generation in conversation and poetry generation. However, automatic generation of news comments is anew, challenging and not well-studied task in NLG. Different from other NLG tasks, this task requires the contextual relevance between comments and news. In addition, we need to generate diversified comments, because different people usually have different opinions on the same news in the real world. In this paper, we propose a gated attention neural network model (GANN) to generate news comments. To address the problem of contextual relevance, we introduce the gated attention mechanism to use news context self-adaptively and selectively. To ensure the diversity of comments, we use random sample and relevance control to generate comments with different topics and degrees of relevance. Moreover, we apply generative adversarial nets to improve GANN. Automatic evaluation with perplexity score reveals that GANN outperforms the existing comment generation methods. Human evaluation proves that the generated news comments are close to human comments.