A Survey on Conversational Recommender Systems

Recommender systems are software applications that help users to find items of interest in situations information overload. Current research often assumes a one-shot interaction paradigm, where the users’ preferences estimated based on past observed behavior and presentation ranked list suggestions is main, one-directional form user interaction. Conversational recommender (CRS) take different approach support richer set interactions. These interactions can, for example, improve preference elicitation process or allow ask questions about recommendations give feedback. The CRS has significantly increased few years. This development mainly due significant progress area natural language processing, emergence new voice-controlled home assistants, use chatbot technology. With this article, we provide detailed survey existing approaches conversational recommendation. We categorize these various dimensions, e.g., terms supported intents knowledge they background. Moreover, discuss technological approaches, review how evaluated, finally identify number gaps deserve more future.