Leveraging Historical Interaction Data for Improving Conversational Recommender System

TL;DR

This paper introduces a novel pre-training method that combines historical interaction data with conversation data to enhance conversational recommender systems, demonstrating improved performance on real-world datasets.

Contribution

The paper proposes a new pre-training approach that fuses item-based and attribute-based preferences using specially designed tasks and a negative sampling method for better CRS performance.

Findings

Effective integration of historical and conversation data improves CRS accuracy.

Designed pre-training tasks enhance information fusion between preference types.

Experimental results show significant performance gains on real-world datasets.

Abstract

Recently, conversational recommender system (CRS) has become an emerging and practical research topic. Most of the existing CRS methods focus on learning effective preference representations for users from conversation data alone. While, we take a new perspective to leverage historical interaction data for improving CRS. For this purpose, we propose a novel pre-training approach to integrating both item-based preference sequence (from historical interaction data) and attribute-based preference sequence (from conversation data) via pre-training methods. We carefully design two pre-training tasks to enhance information fusion between item- and attribute-based preference. To improve the learning performance, we further develop an effective negative sample generator which can produce high-quality negative samples. Experiment results on two real-world datasets have demonstrated the effectiveness of our approach for improving CRS.