Advancing Multi-Party Dialogue Systems with Speaker-ware Contrastive Learning

2025-01-20

AI Summary

Existing multi-party dialogue response generation methods struggle to model speaker-specific stylistic variations and dynamic topic shifts, while relying on complex graph structures and extensive annotated data. To address these limitations, we propose CMR, the first multi-party dialogue generation framework incorporating self-supervised contrastive learning. CMR jointly models stylistic representations and topic evolution via intra-speaker contrastive learning—eliminating the need for explicit graph construction or auxiliary annotations. The model integrates speaker-aware encoding with contextual structural modeling. Evaluated on standard multi-party dialogue benchmarks, CMR achieves significant improvements over state-of-the-art methods: +4.2% in response relevance and +6.8 percentage points in speaker consistency (absolute gain). These results validate the effectiveness of our lightweight, scalable paradigm for joint style-topic modeling in multi-party dialogue generation.

Technology Category

• Machine Learning: Multimodal Learning
• Natural Language Processing: Generation
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

Dialogue response generation has made significant progress, but most research has focused on dyadic dialogue. In contrast, multi-party dialogues involve more participants, each potentially discussing different topics, making the task more complex. Current methods often rely on graph neural networks to model dialogue context, which helps capture the structural dynamics of multi-party conversations. However, these methods are heavily dependent on intricate graph structures and dataset annotations, and they often overlook the distinct speaking styles of participants. To address these challenges, we propose CMR, a Contrastive learning-based Multi-party dialogue Response generation model. CMR uses self-supervised contrastive learning to better distinguish"who says what."Additionally, by comparing speakers within the same conversation, the model captures differences in speaking styles and thematic transitions. To the best of our knowledge, this is the first approach to apply contrastive learning in multi-party dialogue generation. Experimental results show that CMR significantly outperforms state-of-the-art models in multi-party dialogue response tasks.

Problem

• Multi-party Dialogue
• Graph Neural Networks
• Style and Topic Identification

Innovation

• CMR Model
• Contrastive Learning
• Multi-party Dialogue