Conversational recommender systems benefit from persona and knowledge augmentation, as demonstrated by the PEARL dataset.
The author presents PEARL, a dataset addressing limitations in existing conversational recommendation datasets by synthesizing persona- and knowledge-augmented dialogues.