Modeling Multiple User Interests using Hierarchical Knowledge for Conversational Recommender System

Yuka Okuda; Katsuhito Sudoh; Seitaro Shinagawa; Satoshi Nakamura

Modeling Multiple User Interests using Hierarchical Knowledge for Conversational Recommender System

Computation and Language 2023-03-02 v1 Artificial Intelligence Information Retrieval

Authors: Yuka Okuda , Katsuhito Sudoh , Seitaro Shinagawa , Satoshi Nakamura

Abstract

A conversational recommender system (CRS) is a practical application for item recommendation through natural language conversation. Such a system estimates user interests for appropriate personalized recommendations. Users sometimes have various interests in different categories or genres, but existing studies assume a unique user interest that can be covered by closely related items. In this work, we propose to model such multiple user interests in CRS. We investigated its effects in experiments using the ReDial dataset and found that the proposed method can recommend a wider variety of items than that of the baseline CR-Walker.

Keywords

recommendation system recommender system information retrieval

Cite

@article{arxiv.2303.00311,
  title  = {Modeling Multiple User Interests using Hierarchical Knowledge for Conversational Recommender System},
  author = {Yuka Okuda and Katsuhito Sudoh and Seitaro Shinagawa and Satoshi Nakamura},
  journal= {arXiv preprint arXiv:2303.00311},
  year   = {2023}
}

Comments

Accepted as a conference paper at IWSDS 2023

Modeling Multiple User Interests using Hierarchical Knowledge for Conversational Recommender System

Abstract

Keywords

Cite

Comments

Related papers