Open AccessAn improved hybrid recommendation algorithm based on feature preference
Author(s) -
Junjie Xu
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1994/1/012043
Subject(s) - collaborative filtering , computer science , feature (linguistics) , preference , recommender system , data mining , artificial intelligence , algorithm , pattern recognition (psychology) , big data , object (grammar) , machine learning , mathematics , philosophy , linguistics , statistics
In order to solve the problem of data matrix sparsity and cold start in the era of big data. This paper designs a collaborative filtering algorithm, which is an improved hybrid recommendation algorithm based on feature preference analysis. It combines the analysis of user feature preference and item feature, and then uses the traditional collaborative filtering idea to recommend the best scoring object to users. The experimental results show that the accuracy of the proposed algorithm is improved obviously.