Unsupervised Mutual Learning of Discourse Parsing and Topic Segmentation in Dialogue

2024-05-30

AI Summary

To address the reliance on manual annotations and the difficulty in modeling deep interactions between utterance segmentation and topic segmentation in dialogue systems, this paper proposes UMLF—the first unsupervised joint learning framework. Methodologically: (1) it constructs a unified utterance-topic representation space; (2) it introduces discourse-theory-driven local coupling constraints and global topological constraints; and (3) it enables probabilistic co-optimization via pretrained language models. Its key contribution lies in achieving, for the first time, unsupervised mutual learning between the two structural tasks without any labeled data. Empirical evaluation demonstrates consistent superiority over strong baselines across two utterance segmentation and three topic segmentation benchmarks. Furthermore, when adapted to large language models, UMLF significantly improves dialogue state tracking accuracy and contextual response quality.

Technology Category

• Machine Learning: Unsupervised & Self-Supervised Learning
• Natural Language Processing: Conversational AI/Dialog Systems
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

In dialogue systems, discourse plays a crucial role in managing conversational focus and coordinating interactions. It consists of two key structures: rhetorical structure and topic structure. The former captures the logical flow of conversations, while the latter detects transitions between topics. Together, they improve the ability of a dialogue system to track conversation dynamics and generate contextually relevant high-quality responses. These structures are typically identified through discourse parsing and topic segmentation, respectively. However, existing supervised methods rely on costly manual annotations, while unsupervised methods often focus on a single task, overlooking the deep linguistic interplay between rhetorical and topic structures. To address these issues, we first introduce a unified representation that integrates rhetorical and topic structures, ensuring semantic consistency between them. Under the unified representation, we further propose two linguistically grounded hypotheses based on discourse theories: (1) Local Discourse Coupling, where rhetorical cues dynamically enhance topic-aware information flow, and (2) Global Topology Constraint, where topic structure patterns probabilistically constrain rhetorical relation distributions. Building on the unified representation and two hypotheses, we propose an unsupervised mutual learning framework (UMLF) that jointly models rhetorical and topic structures, allowing them to mutually reinforce each other without requiring additional annotations. We evaluate our approach on two rhetorical datasets and three topic segmentation datasets. Experimental results demonstrate that our method surpasses all strong baselines built on pre-trained language models. Furthermore, when applied to LLMs, our framework achieves notable improvements, demonstrating its effectiveness in improving discourse structure modeling.

Problem

• Unsupervised mutual learning
• Discourse parsing
• Topic segmentation

Innovation

• Unsupervised mutual learning framework
• Unified rhetorical-topic representation
• Local-global discourse hypotheses