CluCDD:Contrastive Dialogue Disentanglement via Clustering

TL;DR

CluCDD introduces a contrastive learning-based model for dialogue disentanglement that effectively separates intertwined multi-participant dialogues into distinct sessions, outperforming previous methods on benchmark datasets.

Contribution

The paper proposes CluCDD, a novel contrastive learning approach that improves dialogue disentanglement by aggregating utterances into sessions more accurately than prior clustering methods.

Findings

Achieves state-of-the-art results on Movie Dialogue and IRC datasets.

Effectively separates dialogue sessions with improved clustering performance.

Demonstrates robustness across different multi-party dialogue datasets.

Abstract

A huge number of multi-participant dialogues happen online every day, which leads to difficulty in understanding the nature of dialogue dynamics for both humans and machines. Dialogue disentanglement aims at separating an entangled dialogue into detached sessions, thus increasing the readability of long disordered dialogue. Previous studies mainly focus on message-pair classification and clustering in two-step methods, which cannot guarantee the whole clustering performance in a dialogue. To address this challenge, we propose a simple yet effective model named CluCDD, which aggregates utterances by contrastive learning. More specifically, our model pulls utterances in the same session together and pushes away utterances in different ones. Then a clustering method is adopted to generate predicted clustering labels. Comprehensive experiments conducted on the Movie Dialogue dataset and IRC dataset demonstrate that our model achieves a new state-of-the-art result.