a repository of CRS and MCR.
According to the dataset,I devide the CRS papers into 2 parts. Namely CRS using datasets of pure dialogs e.g. Redial,Inspired.. And MCR to spicfy multi-turn CRS, using more general recommmend dataset e.g. Yelp.. MCR mainly focus on policy selection using RL.
- DeepCPR : Deep Path Reasoning Using Sequence of User-Preferred Attributes for Conversational Recommendation（TKDD 2023）DeepCPR
- Meta-CRS: A Dynamic Meta-Learning Approach for Effective Conversational Recommender System Meta-CRS
- Knowledge Graph Attention Network with Attribute Significance for Personalized Recommendation KG-Personalized Rec