Knowledge-Enhanced Conversational Recommendation via Transformer-Based Sequential Modeling.

The article focuses on the transformer-based sequential conversational recommendation (TSCR) method that models the sequential dependencies of items in conversations to enhance the performance of conversational recommender systems (CRSs). Topics include the deployment of a Cloze task to predict recommended items, the enhancement of the model using knowledge graphs to improve user sequence representation, and the identification of limitations such as the need for sentiment modeling.