Yifei Zhang
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
Research Subjects
Research Interests
Yifei Zhang has a total of 98 co-authors. Here are the top co-authors by number of shared publications:
- Daling Wang
School of Computer Science and Engineering, Northeastern University, Shenyang, China
49 shared publications - Shi Feng
School of Computer Science and Engineering, Northeastern University, Shenyang, China
47 shared publications - Xiaocui Yang
Northeastern University CHINA
12 shared publications - Ge Yu
Northeastern University, Shenyang, China
9 shared publications - Kaisong Song
Tongyi Lab Alibaba Group
8 shared publications - Soujanya Poria
Singapore University of Technology and Design, Singapore, Singapore
5 shared publications
Latest Publications
- Open Access
- 25 August 2026
- Published by Springer Nature in Data Science and Engineering
- p. 1-19
Contrastive Learning with Growing Generated Representations for Inductive Knowledge Graph Embedding
- 25 August 2026
- Published by Springer Nature in Data Science and Engineering
- p. 1-19
Open AccessContrastive Learning with Growing Generated Representations for Inductive Knowledge Graph Embedding
- 23 February 2026
- Published by Springer Nature
- p. 365-380
Adaptive Persona Context Modulation for Personalized Emotional Support Conversation
- 23 February 2026
- Published by Springer Nature
- p. 365-380
Adaptive Persona Context Modulation for Personalized Emotional Support Conversation
- 31 January 2026
- Published by AI Access Foundation in Journal of Artificial Intelligence Research
- Vol. 85
T-COL: Generating Counterfactual Explanations for General User Preferences on Variable Machine Learning Systems
To address the interpretability challenge in machine learning (ML) systems, counterfactual explanations (CEs) have emerged as a promising solution. CEs are unique as they provide workable suggestions to users, instead of explaining why a certain outcome was predicted. The application of CEs encounters two main challenges: general user preferences and variable ML systems. On one hand, user preferences for specific values can vary depending on the task and scenario. On the other hand, the ML systems for verification may change while the CEs are performed. Thus, user preferences tend to be general rather than specific, and CEs need to be adaptable to variable ML models while maintaining robustness even as these models change. Facing these challenges, we propose general user preferences based on insights from psychology and behavioral science, and add the challenge of non-static ML systems as one preference. Moreover, we introduce a novel method, Tree-based Conditions Optional Links (T-COL) for generating CEs adaptable to general user preferences. Moreover, we employ T-COL to enhance the robustness of CEs with specific conditions, making CEs robust even when the ML models are replaced. To assess subjectivity preferences, we define LLM-based autonomous agents to simulate users and align them with real users. Experiments show that T-COL outperforms all baselines in adapting to general user preferences.
- 31 January 2026
- Published by AI Access Foundation in Journal of Artificial Intelligence Research
- Vol. 85
T-COL: Generating Counterfactual Explanations for General User Preferences on Variable Machine Learning Systems
To address the interpretability challenge in machine learning (ML) systems, counterfactual explanations (CEs) have emerged as a promising solution. CEs are unique as they provide workable suggestions to users, instead of explaining why a certain outcome was predicted. The application of CEs encounters two main challenges: general user preferences and variable ML systems. On one hand, user preferences for specific values can vary depending on the task and scenario. On the other hand, the ML systems for verification may change while the CEs are performed. Thus, user preferences tend to be general rather than specific, and CEs need to be adaptable to variable ML models while maintaining robustness even as these models change. Facing these challenges, we propose general user preferences based on insights from psychology and behavioral science, and add the challenge of non-static ML systems as one preference. Moreover, we introduce a novel method, Tree-based Conditions Optional Links (T-COL) for generating CEs adaptable to general user preferences. Moreover, we employ T-COL to enhance the robustness of CEs with specific conditions, making CEs robust even when the ML models are replaced. To assess subjectivity preferences, we define LLM-based autonomous agents to simulate users and align them with real users. Experiments show that T-COL outperforms all baselines in adapting to general user preferences.
- Top Cited Papers
- 25 January 2026
- Published by Elsevier in Knowledge-Based Systems
- Vol. 337
Affective computing in the era of large language models: A survey from the NLP perspective
- 25 January 2026
- Published by Elsevier in Knowledge-Based Systems
- Vol. 337
Top Cited PapersAffective computing in the era of large language models: A survey from the NLP perspective
- 2 January 2026
- Published by Springer Nature
- p. 209-224
Enhancing Zero-Shot Emotion Perception in Conversation Through the Internal-to-External Chain-of-Thought
- 2 January 2026
- Published by Springer Nature
- p. 209-224
Enhancing Zero-Shot Emotion Perception in Conversation Through the Internal-to-External Chain-of-Thought
- Open Access
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 14122-14129
A Two-Stage Full Fine-Tuning and LLM Post-processing Framework for MCABSA
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 14122-14129
Open AccessA Two-Stage Full Fine-Tuning and LLM Post-processing Framework for MCABSA
- Open Access
MEKiT: Multi-source Heterogeneous Knowledge Injection Method via Instruction Tuning for Emotion-Cause Pair Extraction
Although large language models (LLMs) excel in text comprehension and generation, their performance on the Emotion-Cause Pair Extraction (ECPE) task, which requires reasoning ability, is often underperform smaller language model. The main reason is the lack of auxiliary knowledge, which limits LLMs' ability to effectively perceive emotions and reason causes. To address this issue, we propose a novel \textbf{M}ulti-source h\textbf{E}terogeneous \textbf{K}nowledge \textbf{i}njection me\textbf{T}hod, MEKiT, which integrates heterogeneous internal emotional knowledge and external causal knowledge. Specifically, for these two distinct aspects and structures of knowledge, we apply the approaches of incorporating instruction templates and mixing data for instruction-tuning, which respectively facilitate LLMs in more comprehensively identifying emotion and accurately reasoning causes. Experimental results demonstrate that MEKiT provides a more effective and adaptable solution for the ECPE task, exhibiting an absolute performance advantage over compared baselines and dramatically improving the performance of LLMs on the ECPE task.
Open AccessMEKiT: Multi-source Heterogeneous Knowledge Injection Method via Instruction Tuning for Emotion-Cause Pair Extraction
Although large language models (LLMs) excel in text comprehension and generation, their performance on the Emotion-Cause Pair Extraction (ECPE) task, which requires reasoning ability, is often underperform smaller language model. The main reason is the lack of auxiliary knowledge, which limits LLMs' ability to effectively perceive emotions and reason causes. To address this issue, we propose a novel \textbf{M}ulti-source h\textbf{E}terogeneous \textbf{K}nowledge \textbf{i}njection me\textbf{T}hod, MEKiT, which integrates heterogeneous internal emotional knowledge and external causal knowledge. Specifically, for these two distinct aspects and structures of knowledge, we apply the approaches of incorporating instruction templates and mixing data for instruction-tuning, which respectively facilitate LLMs in more comprehensively identifying emotion and accurately reasoning causes. Experimental results demonstrate that MEKiT provides a more effective and adaptable solution for the ECPE task, exhibiting an absolute performance advantage over compared baselines and dramatically improving the performance of LLMs on the ECPE task.
- Open Access
- 30 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 1394-1403
Generative Emotion Cause Explanation in Multimodal Conversations
- 30 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 1394-1403
Open AccessGenerative Emotion Cause Explanation in Multimodal Conversations
- Open Access
Enhancing LLM-based Recommendation through Semantic-Aligned Collaborative Knowledge
Large Language Models (LLMs) demonstrate remarkable capabilities in leveraging comprehensive world knowledge and sophisticated reasoning mechanisms for recommendation tasks. However, a notable limitation lies in their inability to effectively model sparse identifiers (e.g., user and item IDs), unlike conventional collaborative filtering models (Collabs.), thus hindering LLM to learn distinctive user-item representations and creating a performance bottleneck. Prior studies indicate that integrating collaborative knowledge from Collabs. into LLMs can mitigate the above limitations and enhance their recommendation performance. Nevertheless, the significant discrepancy in knowledge distribution and semantic space between LLMs and Collab. presents substantial challenges for effective knowledge transfer. To tackle these challenges, we propose a novel framework, SeLLa-Rec, which focuses on achieving alignment between the semantic spaces of Collabs. and LLMs. This alignment fosters effective knowledge fusion, mitigating the influence of discriminative noise and facilitating the deep integration of knowledge from diverse models. Specifically, three special tokens with collaborative knowledge are embedded into the LLM's semantic space through a hybrid projection layer and integrated into task-specific prompts to guide the recommendation process. Experiments conducted on two public benchmark datasets (MovieLens-1M and Amazon Book) demonstrate that SeLLa-Rec achieves state-of-the-art performance.
Open AccessEnhancing LLM-based Recommendation through Semantic-Aligned Collaborative Knowledge
Large Language Models (LLMs) demonstrate remarkable capabilities in leveraging comprehensive world knowledge and sophisticated reasoning mechanisms for recommendation tasks. However, a notable limitation lies in their inability to effectively model sparse identifiers (e.g., user and item IDs), unlike conventional collaborative filtering models (Collabs.), thus hindering LLM to learn distinctive user-item representations and creating a performance bottleneck. Prior studies indicate that integrating collaborative knowledge from Collabs. into LLMs can mitigate the above limitations and enhance their recommendation performance. Nevertheless, the significant discrepancy in knowledge distribution and semantic space between LLMs and Collab. presents substantial challenges for effective knowledge transfer. To tackle these challenges, we propose a novel framework, SeLLa-Rec, which focuses on achieving alignment between the semantic spaces of Collabs. and LLMs. This alignment fosters effective knowledge fusion, mitigating the influence of discriminative noise and facilitating the deep integration of knowledge from diverse models. Specifically, three special tokens with collaborative knowledge are embedded into the LLM's semantic space through a hybrid projection layer and integrated into task-specific prompts to guide the recommendation process. Experiments conducted on two public benchmark datasets (MovieLens-1M and Amazon Book) demonstrate that SeLLa-Rec achieves state-of-the-art performance.
- 22 February 2025
- Published by Springer Nature in International Journal of Machine Learning and Cybernetics
- p. 1-18
AC-EIC: addressee-centered emotion inference in conversations
- 22 February 2025
- Published by Springer Nature in International Journal of Machine Learning and Cybernetics
- p. 1-18
AC-EIC: addressee-centered emotion inference in conversations