Journal (ESWA)
2026
From Consensus Theory to LLM Agents: Practical Consensus-Building for Multi-Issue Negotiation
Yihan Dong, Takayuki Ito
Expert Systems with Applications, 2026.
Academic Specialist
Yihan Dong
Graduate School of Engineering, The University of Tokyo
Postdoctoral researcher focused on Multi-Agent Systems, Game Theory, and Large Language Models, with a particular emphasis on formal consensus-reaching mechanisms and explainable decision-making.
Multi-Agent System
Game Theory
Consensus-Reaching Process
Group Decision-Making
Large Language Model
Research Framework
Consensus-Reaching Process Support with LLM-Based Multi-Agent Systems
My research focuses on establishing a principled framework for LLM-based Multi-Agent Systems (MAS) to support credible and interpretable consensus-reaching.
01
Structured Facilitation
Demonstrating practical online deliberation by embedding LLM-based agents into explicit multi-phase workflows.
02
Explainable Credibility Judgment
Making the results of LLM-based MAS decision-making explainable by designing an aggregation mechanism rather than solely depending on "LLM-as-a-judge".
03
Fuzzy Preference Elicitation & Analysis
Bridging linguistic preferences and formal analysis through fuzzy preference elicitation and tolerance-aware judgment.
04
Cross-Paradigm Evaluation
Formalising the multi-issue negotiation framework for LLM-based MAS to evaluate stability, cost, and fairness across mechanism-governed and prompt-based agents.
Publications
Refereed journals, conferences, and symposia
Refereed Journal Articles
Journal (Sci Rep)
2026
Multi-agent systems and credibility-based advanced scoring mechanism in fact-checking
Yihan Dong, Takayuki Ito
Scientific Reports, 2026.
Journal (IEICE)
2025
A Linguistic Negotiation for Consensus-Making among LLM Agents with Fuzzy Utility Functions
Yihan Dong, Takayuki Ito
IEICE Transactions on Information and Systems, 2025.
Journal (IEICE)
2023
An Automated Multi-Phase Facilitation Agent Based on LLM
Yihan Dong, Shiyao Ding, Takayuki Ito
IEICE Transactions on Information and Systems, 2023.
International Conferences & Consortiums
AAMAS DC
2024
The Multi-agent System based on LLM for Online Discussions
Yihan Dong
AAMAS Doctoral Consortium, 2024.
IEEE ICA
2023
Human Discussion Support through Multiagent System: A Multi-Phase Facilitation Agent based on LLM
Dong, Y., & Ito, T.
The IEEE International Conference on Agentic AI (IEEE ICA) poster, 2023.
Domestic Conferences & Symposia (Japan)
Multi-agent Fact-checking based on Large Language Models
Dong, Y., & Ito, T. • IPSJ 86th National Convention, 2024.
Fact-Checking Method for Multi-Agent Systems Based on Large Language Models
Dong, Y., & Ito, T. • JSAI 38th Annual Conference, 2024.
A Case Study of the Automated Facilitation Agent Based on LLM
Dong, Y., & Ito, T. • Joint Agent Workshops and Symposium (JAWS), 2023.
Comparison of the Accuracy of Classifying English Sentences According to IBIS Labels
Dong, Y., & Ito, T. • IPSJ 85th National Convention, 2023.
An Implementation of An Automated Facilitation Agent Promoting Inclusive Discussion
Dong, Y., & Ito, T. • JSAI 37th Annual Conference, 2023.
Career
2026/10 – Present
Academic Specialist
The University of Tokyo
Graduate School of Engineering • Tokyo
2026/08 – 2026/09
Research Internship
NEC Research & Development
Yokohama, Kanagawa
Education
2022/10 – 2026/09
Ph.D. in Informatics
Kyoto University
Supervisor: Prof. Takayuki Ito • Kyoto, Japan
2020/01 – 2021/12
Master of Engineering
The University of Melbourne
Melbourne, Victoria, Australia
2015/09 – 2019/06
Bachelor of Software Engineering
Northeastern University
Shenyang, Liaoning, China
Honours & Awards
Outstanding Research Award
18th ICT Innovation, Kyoto University (2024/02/21)
"LLM-based Multi-agent Fact-checking System for Online Discussions"
Research Funding
Kyoto University Science and Technology Innovation Creation Fellowship
Period: 2023/04/01 – 2025/09/30
Skills & Languages
Programming:
Python (Proficient)
Languages:
English (Business), Japanese (Conversational), Mandarin (Native)