A conversation recommender system (CRS) aims to provide high-quality recommendations to users in fewer conversation turns. Existing studies often rely on knowledge graphs and entities mentioned in dialogues to enhance the representation of entity information. However, they fail to thoroughly explore users’ emotional tendencies toward entities or effectively differentiate the varying impacts of different entities on user preferences. Recently, Liu et.al. proposed an innovative Sentimentally Enhanced Conversation Recommender system (SECR). In this presentation, the SECR method will be explained and described how this method captured users' emotional inclinations toward entities.
