Personalized Topic Selection Model for Topic-Grounded Dialogue

TL;DR

This paper introduces PETD, a personalized topic selection model for topic-grounded dialogue that effectively integrates multiple sources of side information to improve relevance and engagement in conversations.

Contribution

The paper proposes a novel PETD model that selectively combines side information like topics and personas, using contrastive learning to enhance topic relevance and dialogue coherence.

Findings

Outperforms state-of-the-art baselines in multiple metrics

Generates more engaging and diverse responses

Effectively filters irrelevant personas using contrastive learning

Abstract

Recently, the topic-grounded dialogue (TGD) system has become increasingly popular as its powerful capability to actively guide users to accomplish specific tasks through topic-guided conversations. Most existing works utilize side information (\eg topics or personas) in isolation to enhance the topic selection ability. However, due to disregarding the noise within these auxiliary information sources and their mutual influence, current models tend to predict user-uninteresting and contextually irrelevant topics. To build user-engaging and coherent dialogue agent, we propose a \textbf{P}ersonalized topic s\textbf{E}lection model for \textbf{T}opic-grounded \textbf{D}ialogue, named \textbf{PETD}, which takes account of the interaction of side information to selectively aggregate such information for more accurately predicting subsequent topics. Specifically, we evaluate the correlation between global topics and personas and selectively incorporate the global topics aligned with user personas. Furthermore, we propose a contrastive learning based persona selector to filter out irrelevant personas under the constraint of lacking pertinent persona annotations. Throughout the selection and generation, diverse relevant side information is considered. Extensive experiments demonstrate that our proposed method can generate engaging and diverse responses, outperforming state-of-the-art baselines across various evaluation metrics.