Search papers, labs, and topics across Lattice.
This paper introduces HyCoRec, a novel approach that leverages hypergraph structures to enhance multi-preference learning in conversational recommendation systems, specifically targeting the Matthew effect where popular items dominate recommendations. By integrating multiple preference aspects鈥攊tem, entity, word, review, and knowledge鈥擧yCoRec effectively generates contextually relevant responses and improves item prediction accuracy over time. Experimental results on two benchmark datasets demonstrate that HyCoRec not only achieves state-of-the-art performance but also significantly mitigates the Matthew effect in user interactions.
Alleviating the Matthew effect in conversational recommendations could revolutionize how users discover less popular items, enhancing diversity in recommendations.
The Matthew effect is a notorious issue in Recommender Systems (RSs), \emph{i.e.}, the rich get richer and the poor get poorer, wherein popular items are overexposed while less popular ones are regularly ignored. Most methods examine Matthew effect in static or nearly-static recommendation scenarios. However, the Matthew effect will be increasingly amplified when the user interacts with the system over time. To address these issues, we propose a novel paradigm, Hypergraph-Enhanced Multi-Preference Learning for Alleviating Matthew Effect in Conversational Recommendation (HyCoRec), which aims to alleviate the Matthew effect in conversational recommendation. Concretely, HyCoRec devotes to alleviate the Matthew effect by learning multi-aspect preferences, \emph{i.e.}, item-, entity-, word-, review-, and knowledge-aspect preferences, to effectively generate responses in the conversational task and accurately predict items in the recommendation task when the user chats with the system over time. Extensive experiments conducted on two benchmarks validate that HyCoRec achieves new state-of-the-art performance and the superior of alleviating Matthew effect. Our code is available at https://github.com/zysensmile/HyCoRec.
