Towards Knowledge-Based Recommender Dialog System

TL;DR

The paper presents KBRD, an end-to-end Knowledge-Based Recommender Dialog System that tightly couples a recommender with a knowledge-grounded dialog generator to leverage user preferences and provide recommendation-aware vocabulary. It demonstrates mutual benefits: the dialog component gains from knowledge-grounded information, while the recommender injects vocabulary bias into the dialog, improving overall performance. Experimental results show substantial gains over baselines in both dialog quality and recommendation accuracy, supported by analyses of mutual reinforcement and knowledge contribution. This approach advances practical, knowledge-infused dialogue for personalized recommendations with potential impact on user-centric interfaces in domains like e-commerce and content platforms.

Abstract

In this paper, we propose a novel end-to-end framework called KBRD, which stands for Knowledge-Based Recommender Dialog System. It integrates the recommender system and the dialog generation system. The dialog system can enhance the performance of the recommendation system by introducing knowledge-grounded information about users' preferences, and the recommender system can improve that of the dialog generation system by providing recommendation-aware vocabulary bias. Experimental results demonstrate that our proposed model has significant advantages over the baselines in both the evaluation of dialog generation and recommendation. A series of analyses show that the two systems can bring mutual benefits to each other, and the introduced knowledge contributes to both their performances.