Topic-Aware Contrastive Learning for Abstractive Dialogue Summarization

TL;DR

This paper introduces topic-aware contrastive learning objectives to improve abstractive dialogue summarization by capturing topic changes and handling scattered information, leading to state-of-the-art results.

Contribution

It proposes two novel contrastive learning objectives specifically designed for dialogue summarization, addressing topic variation and information scattering challenges.

Findings

Significant performance improvement over strong baselines.

Achieves new state-of-the-art results on benchmark datasets.

Contrastive objectives effectively model topic changes in dialogues.

Abstract

Unlike well-structured text, such as news reports and encyclopedia articles, dialogue content often comes from two or more interlocutors, exchanging information with each other. In such a scenario, the topic of a conversation can vary upon progression and the key information for a certain topic is often scattered across multiple utterances of different speakers, which poses challenges to abstractly summarize dialogues. To capture the various topic information of a conversation and outline salient facts for the captured topics, this work proposes two topic-aware contrastive learning objectives, namely coherence detection and sub-summary generation objectives, which are expected to implicitly model the topic change and handle information scattering challenges for the dialogue summarization task. The proposed contrastive objectives are framed as auxiliary tasks for the primary dialogue summarization task, united via an alternative parameter updating strategy. Extensive experiments on benchmark datasets demonstrate that the proposed simple method significantly outperforms strong baselines and achieves new state-of-the-art performance. The code and trained models are publicly available via \href{https://github.com/Junpliu/ConDigSum}{https://github.com/Junpliu/ConDigSum}.