User-Centric Conversational Recommendation with Multi-Aspect User Modeling (UCCR)

[KDD23] Official PyTorch implementation for "Improving Conversational Recommendation Systems via Counterfactual Data Simulation".

Official PyTorch implementation for "Multi-grained Hypergraph Interest Modeling for Conversational Recommendation".

[Official Code] Improving Conversational Recommendation Systems via Bias Analysis and Language-Model-Enhanced Data Augmentation

Aspect Based ReDial(AB-ReDial): Is a subset data from ReDial annotated on six dialogue aspects and overall user satisfaction at the turn and dialogue levels with the following aspects; relevance, interestingness, understanding, task completion, interest arousal, and efficiency

Conversational recommender system research

ECR-RR contains the complete implementation of an inference-time reranking system for empathetic conversational recommendation (ECR: https://github.com/zxd-octopus/ECR), featuring a RoBERTa-based critic for response quality assessment and NDCG-balanced evaluation.