Who Responded to Whom: The Joint Effects of Latent Topics and Discourse in Conversation Structure

TL;DR

This paper presents a model that leverages latent topics and discourse roles to accurately identify response relations in online conversations, improving over previous methods and providing insights into conversational dynamics.

Contribution

The authors introduce a novel model that jointly learns latent topics and discourse roles to predict who responds to whom in conversations, addressing limitations of prior work.

Findings

Achieves 79 MRR on Chinese customer service dialogues, outperforming previous state-of-the-art of 73.

Effectively captures the influence of latent topics and discourse roles on response prediction.

Provides interpretability into how topics and discourse shape conversational interactions.

Abstract

Numerous online conversations are produced on a daily basis, resulting in a pressing need to conversation understanding. As a basis to structure a discussion, we identify the responding relations in the conversation discourse, which link response utterances to their initiations. To figure out who responded to whom, here we explore how the consistency of topic contents and dependency of discourse roles indicate such interactions, whereas most prior work ignore the effects of latent factors underlying word occurrences. We propose a model to learn latent topics and discourse in word distributions, and predict pairwise initiation-response links via exploiting topic consistency and discourse dependency. Experimental results on both English and Chinese conversations show that our model significantly outperforms the previous state of the arts, such as 79 vs. 73 MRR on Chinese customer service dialogues. We further probe into our outputs and shed light on how topics and discourse indicate conversational user interactions.