MFSR: A novel multi-level fuzzy similarity measure for recommender systems
作者：
Highlights：
• A suggested fuzzy similarity measure based on Popularity and Significance.
• Improving uncertainty and similarity calculation in collaborative filtering.
• Using the Multi-Level structure to decrease the complexity.
• Improvement of Mean Absolute Error and accuracy of the recommendations.
• Results show better uncertainty handling in collaborative filtering than former studies.
摘要
•A suggested fuzzy similarity measure based on Popularity and Significance.•Improving uncertainty and similarity calculation in collaborative filtering.•Using the Multi-Level structure to decrease the complexity.•Improvement of Mean Absolute Error and accuracy of the recommendations.•Results show better uncertainty handling in collaborative filtering than former studies.
论文关键词：Recommender system,Collaborative filtering,Similarity measure,Fuzzy logic,Multi-level similarity,MFSR
论文评审过程：Received 17 May 2020, Revised 21 February 2021, Accepted 27 March 2021, Available online 31 March 2021, Version of Record 16 April 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2021.114969
