Discriminate2Rec: Negation-based dynamic discriminative interest-based preference learning for semantics-aware content-based recommendation
作者：
Highlights：
• A three-stage preference learning approach to content-based recommender systems.
• Semantical and temporal attribute-level coherence of user profile are improved.
• User profile is learned based on discriminative user-attribute interest weights.
• The discrimination between items’ attributes improves the recommendation accuracy.
• Introducing a negation-based profile modelling method for accurate recommendation.
摘要
•A three-stage preference learning approach to content-based recommender systems.•Semantical and temporal attribute-level coherence of user profile are improved.•User profile is learned based on discriminative user-attribute interest weights.•The discrimination between items’ attributes improves the recommendation accuracy.•Introducing a negation-based profile modelling method for accurate recommendation.
论文关键词：Content-based filtering,Semantics-aware recommender systems,Temporal dynamics,User discriminative interests,Preference learning,Profile coherence
论文评审过程：Received 22 July 2021, Revised 13 March 2022, Accepted 24 March 2022, Available online 26 March 2022, Version of Record 1 April 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2022.116988
