Dialogue Summarization with Emotion Dynamics Using Topic- and Participant-Centric Decomposition

2026-07-16

AI Summary

This work addresses the limitations of existing dialogue summarization methods, which often overlook multi-participant interactions and emotional dynamics, thereby failing to capture the co-evolution of semantics and affect. To overcome this, the authors propose a novel framework that decomposes dialogues along dual dimensions—topics and participants—and integrates multimodal inputs with emotion inference to jointly model semantic content and emotional trajectories. Key innovations include the explicit incorporation of emotional dynamics into summarization for the first time, a dual-path summary generation mechanism, and a new evaluation metric tailored to emotional trajectory fidelity. Experimental results demonstrate that a small language model based on an enhanced hierarchical Chain-of-Agents architecture can generate high-quality, emotion-aware summaries even without explicit emotion labels.

Abstract

Existing text summarization research has focused much on monologic information (e.g., newspaper articles, reports) without accounting for the interaction between speakers or authors. In contrast, dialogues are a rich communication channel where multiple participants conduct back and forth exchanges to construct meaning. We propose a dialogue summarization framework that explicitly models both semantic and emotion dynamics using multimodal dialogue inputs, built on an adapted hierarchical Chain-of-Agents approach. We decompose dialogues from two perspectives: (1) topic segments based on the utterances of all participants, and (2) participant-specific utterance segments. These are used to generate corresponding summaries while incorporating automatically inferred emotions. Topic- and participant-level summaries are aggregated into a dialogue summary capturing semantic content and emotion trajectories. To evaluate beyond content accuracy, we introduce emotion trajectory metrics measuring how well summaries preserve emotional flow. Experiments with small language models on multimodal dialogue datasets show that our framework produces summaries with both semantic and emotion content. Further experiments on explicit emotion label availability highlight the efficacy of our proposed methodology and the opportunities in dialogue analysis using language models.

Problem

• dialogue summarization
• emotion dynamics
• participant interaction
• topic segmentation
• multimodal dialogue

Innovation

• dialogue summarization
• emotion dynamics
• topic-participant decomposition
• Chain-of-Agents
• multimodal dialogue