Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

MIND: Multi-agent inference for negotiation dialogue in travel planning

Hunmin Do, Taejun Yoon, Kiyong Jung

Abstract

While Multi-Agent Debate (MAD) research has advanced, its efficacy in coordinating complex stakeholder interests such as travel planning remains largely unexplored. To bridge this gap, we propose MIND (Multi-agent Inference for Negotiation Dialogue), a framework designed to simulate realistic consensus-building among travelers with heterogeneous preferences. Grounded in the Theory of Mind (ToM), MIND introduces a Strategic Appraisal phase that infers opponent willingness (w) from linguistic nuances with 90.2% accuracy. Experimental results demonstrate that MIND outperforms traditional MAD frameworks, achieving a 20.5% improvement in High-w Hit and a 30.7% increase in Debate Hit-Rate, effectively prioritizing high-stakes constraints. Furthermore, qualitative evaluations via LLM-as-a-Judge confirm that MIND surpasses baselines in Rationality (68.8%) and Fluency (72.4%), securing an overall win rate of 68.3%. These findings validate that MIND effectively models human negotiation dynamics to derive persuasive consensus.

MIND: Multi-agent inference for negotiation dialogue in travel planning

Abstract

While Multi-Agent Debate (MAD) research has advanced, its efficacy in coordinating complex stakeholder interests such as travel planning remains largely unexplored. To bridge this gap, we propose MIND (Multi-agent Inference for Negotiation Dialogue), a framework designed to simulate realistic consensus-building among travelers with heterogeneous preferences. Grounded in the Theory of Mind (ToM), MIND introduces a Strategic Appraisal phase that infers opponent willingness (w) from linguistic nuances with 90.2% accuracy. Experimental results demonstrate that MIND outperforms traditional MAD frameworks, achieving a 20.5% improvement in High-w Hit and a 30.7% increase in Debate Hit-Rate, effectively prioritizing high-stakes constraints. Furthermore, qualitative evaluations via LLM-as-a-Judge confirm that MIND surpasses baselines in Rationality (68.8%) and Fluency (72.4%), securing an overall win rate of 68.3%. These findings validate that MIND effectively models human negotiation dynamics to derive persuasive consensus.
Paper Structure (28 sections, 2 figures, 21 tables)

This paper contains 28 sections, 2 figures, 21 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. LLM-based Travel Planning
  4. Multi-Agent Debate (MAD) and Social Cognition
  5. Methodology
  6. Multi-Persona Data Augmentation
  7. MIND (Multi-agent Inference for Negotiation Dialogue)
  8. Experiments and Results
  9. Experimental Setup & Metrics
  10. ToM Inference Accuracy.
  11. item Qualitative Evaluation (LLM-as a-Judge).
  12. Results and Analysis
  13. Quantitative Performance & Strategic Trade-off.
  14. Scalability Analysis.
  15. Accuracy of ToM Inference.
  16. ...and 13 more sections

Figures (2)

  • Figure 1: Overview of the MIND Framework. (Left) Persona profiles are augmented with Willingness ($w$) values. (Center) Agents engage in the MIND Loop, inferring opponents' hidden $w$ and adjusting strategies. (Right) The process yields a consensus travel plan.
  • Figure 2: The structured system prompt used to augment persona preferences.