Ruizhang Huang
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
Ruizhang Huang has a total of 56 co-authors. Here are the top co-authors by number of shared publications:
- Yanping Chen
State Key Laboratory of Public Big Data, Guizhou University, Guiyang, China
16 shared publications - Yongbin Qin
Text Computing & Cognitive Intelligence Engineering Research Center of National Education Ministry, College of Computer Science and Technology, Guizhou University, Guiyang, China
13 shared publications - Chuan Lin
College of Computer Science & Technology, Guizhou University, Guiyang, China
5 shared publications - Ruina Bai
Guizhou Univ, Coll Comp Sci & Technol, Text Comp & Cognit Intelligence Engn Res Ctr, State Key Lab Publ Big Data,Natl Educ Minist, Guiyang, Guizhou, Peoples R China
5 shared publications Yanping Chen
Engineering Research Center of National Education Ministry College of Computer Science and Technology Text Computing & Cognitive Intelligence Guizhou University Guiyang China
5 shared publications- Jingjing Xue
Text Computing & Cognitive Intelligence Engineering Research Center of National Education Ministry, State Key Laboratory of Public Big Data, College of Computer Science and Technology, Guizhou University, Guiyang, China
4 shared publications
Latest Publications
- 13 April 2026
- Published by Elsevier in Information Processing & Management
- Vol. 63 (7)
LLM-guided multi-view representation learning for legal case retrieval via structured event chains and role-aware aggregation
- 13 April 2026
- Published by Elsevier in Information Processing & Management
- Vol. 63 (7)
LLM-guided multi-view representation learning for legal case retrieval via structured event chains and role-aware aggregation
- 20 March 2026
- Published by Elsevier in Expert Systems with Applications
- Vol. 319
A causality-aware and hallucination-mitigation framework for faithful news summarization
- 20 March 2026
- Published by Elsevier in Expert Systems with Applications
- Vol. 319
A causality-aware and hallucination-mitigation framework for faithful news summarization
- 10 February 2026
- Published by Elsevier in Pattern Recognition
- Vol. 177
Synergistic representation learning via dual sample interactions for multi-view clustering
- 10 February 2026
- Published by Elsevier in Pattern Recognition
- Vol. 177
Synergistic representation learning via dual sample interactions for multi-view clustering
- 3 February 2026
- Published by Elsevier in Information Processing & Management
- Vol. 63 (5)
DUIC: User-descriptive intention guided clustering for personalized and understandable document partitions
- 3 February 2026
- Published by Elsevier in Information Processing & Management
- Vol. 63 (5)
DUIC: User-descriptive intention guided clustering for personalized and understandable document partitions
- 11 January 2026
- Published by Elsevier in Information Processing & Management
- Vol. 63 (4)
PMDS: progressive multi-document summarization with iterative summary integration
- 11 January 2026
- Published by Elsevier in Information Processing & Management
- Vol. 63 (4)
PMDS: progressive multi-document summarization with iterative summary integration
- 3 December 2025
- Published by Elsevier in Information Fusion
- Vol. 129
Partitioning by preference: Discovering personalized document clusters via user-descriptive clustering intentions
- 3 December 2025
- Published by Elsevier in Information Fusion
- Vol. 129
Partitioning by preference: Discovering personalized document clusters via user-descriptive clustering intentions
- Open Access
A Reasoning Paradigm for Named Entity Recognition
Generative LLMs typically improve Named Entity Recognition (NER) performance through instruction tuning. They excel at generating entities by semantic pattern matching but lack an explicit, verifiable reasoning mechanism. This "cognitive shortcutting" leads to suboptimal performance and brittle generalization, especially in zero-shot and lowresource scenarios where reasoning from limited contextual cues is crucial. To address this issue, a reasoning framework is proposed for NER, which shifts the extraction paradigm from implicit pattern matching to explicit reasoning. This framework consists of three stages: Chain of Thought (CoT) generation, CoT tuning, and reasoning enhancement. First, a dataset annotated with NER-oriented CoTs is generated, which contain task-relevant reasoning chains. Then, they are used to tune the NER model to generate coherent rationales before deriving the final answer. Finally, a reasoning enhancement stage is implemented to optimize the reasoning process using a comprehensive reward signal. This stage ensures explicit and verifiable extractions. Experiments show that ReasoningNER demonstrates impressive cognitive ability in the NER task, achieving competitive performance. In zero-shot settings, it achieves state-of-the-art (SOTA) performance, outperforming GPT-4 by 12.3 percentage points on the F1 score. Analytical results also demonstrate its great potential to advance research in reasoningoriented information extraction. Our codes are available at https://github.com/HuiResearch/ReasoningIE.
Open AccessA Reasoning Paradigm for Named Entity Recognition
Generative LLMs typically improve Named Entity Recognition (NER) performance through instruction tuning. They excel at generating entities by semantic pattern matching but lack an explicit, verifiable reasoning mechanism. This "cognitive shortcutting" leads to suboptimal performance and brittle generalization, especially in zero-shot and lowresource scenarios where reasoning from limited contextual cues is crucial. To address this issue, a reasoning framework is proposed for NER, which shifts the extraction paradigm from implicit pattern matching to explicit reasoning. This framework consists of three stages: Chain of Thought (CoT) generation, CoT tuning, and reasoning enhancement. First, a dataset annotated with NER-oriented CoTs is generated, which contain task-relevant reasoning chains. Then, they are used to tune the NER model to generate coherent rationales before deriving the final answer. Finally, a reasoning enhancement stage is implemented to optimize the reasoning process using a comprehensive reward signal. This stage ensures explicit and verifiable extractions. Experiments show that ReasoningNER demonstrates impressive cognitive ability in the NER task, achieving competitive performance. In zero-shot settings, it achieves state-of-the-art (SOTA) performance, outperforming GPT-4 by 12.3 percentage points on the F1 score. Analytical results also demonstrate its great potential to advance research in reasoningoriented information extraction. Our codes are available at https://github.com/HuiResearch/ReasoningIE.
- 10 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 27 (15209210) , 8185-8198
Optimizing Adaptive Video Streaming: Offline Reinforcement Learning and Meta-Learning in Diverse Networks
Recent years have seen the optimization of quality of experience (QoE) through learning adaptive bitrate (ABR) algorithms from internet video streams. However, the complex nature of the real-world Internet, characterized by heavy-tailed behavior, diversity, and unpredictability, hinder the effective learning of off-the-shelf reinforcement learning (RL)-based ABR algorithms. As a result, existing methods inevitably fail to achieve optimal performance under various network conditions and user QoE objectives. We propose Fortuna, a novel offline meta RL ABR algorithm that can effectively learn from these heavy-tailed internet data features and become more practical. Fortuna is primarily divided into two phases. In the offline phase, Fortuna utilizes diverse offline data for learning to reduce the costly online RL interaction expense, while in the online phase, we gradually increase video streaming sessions complexity through curriculum learning to quickly adapt to specific network conditions. Fortuna then utilizes meta-learning to optimize ABR policies and enhance generalization. Additionally, to better learn network features, Fortuna further optimizes QoE by learning low-level TCP congestion control information. Experimental results from trace-driven and real-world scenarios demonstrate that Fortuna enhances learning efficiency by more than 7.5%-4 ×, reduces stall time by 4.6%-14.2%, and generalizes to different network conditions and video streams.
- 10 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 27 (15209210) , 8185-8198
Optimizing Adaptive Video Streaming: Offline Reinforcement Learning and Meta-Learning in Diverse Networks
Recent years have seen the optimization of quality of experience (QoE) through learning adaptive bitrate (ABR) algorithms from internet video streams. However, the complex nature of the real-world Internet, characterized by heavy-tailed behavior, diversity, and unpredictability, hinder the effective learning of off-the-shelf reinforcement learning (RL)-based ABR algorithms. As a result, existing methods inevitably fail to achieve optimal performance under various network conditions and user QoE objectives. We propose Fortuna, a novel offline meta RL ABR algorithm that can effectively learn from these heavy-tailed internet data features and become more practical. Fortuna is primarily divided into two phases. In the offline phase, Fortuna utilizes diverse offline data for learning to reduce the costly online RL interaction expense, while in the online phase, we gradually increase video streaming sessions complexity through curriculum learning to quickly adapt to specific network conditions. Fortuna then utilizes meta-learning to optimize ABR policies and enhance generalization. Additionally, to better learn network features, Fortuna further optimizes QoE by learning low-level TCP congestion control information. Experimental results from trace-driven and real-world scenarios demonstrate that Fortuna enhances learning efficiency by more than 7.5%-4 ×, reduces stall time by 4.6%-14.2%, and generalizes to different network conditions and video streams.
- 6 January 2025
- Published by Elsevier in Neural Networks
- Vol. 184, 107117
A discrete convolutional network for entity relation extraction
- 6 January 2025
- Published by Elsevier in Neural Networks
- Vol. 184, 107117
A discrete convolutional network for entity relation extraction
- Open Access
- 1 January 2025
- Published by Association for Computational Linguistics (ACL)
- p. 21673-21686
Exclusion of Thought: Mitigating Cognitive Load in Large Language Models for Enhanced Reasoning in Multiple-Choice Tasks
Qihang Fu, Yongbin Qin, Ruizhang Huang, Yanping Chen, Yulin Zhou, Lintao Long. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.
- 1 January 2025
- Published by Association for Computational Linguistics (ACL)
- p. 21673-21686
Open AccessExclusion of Thought: Mitigating Cognitive Load in Large Language Models for Enhanced Reasoning in Multiple-Choice Tasks
Qihang Fu, Yongbin Qin, Ruizhang Huang, Yanping Chen, Yulin Zhou, Lintao Long. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.