Personalized recommendation by matrix co-factorization with tags and time information

作者：

Highlights：

• Free tags and temporal information are adopted to obtain precise user preferences.

• Matrices co-factorization method is proposed to the joint the extracted information.

• Experiments verify that integrating tag information is helpful to relieve overfitting.

• The proposed co-SVD model outperforms baseline methods significantly.

摘要

•Free tags and temporal information are adopted to obtain precise user preferences.•Matrices co-factorization method is proposed to the joint the extracted information.•Experiments verify that integrating tag information is helpful to relieve overfitting.•The proposed co-SVD model outperforms baseline methods significantly.

论文关键词：Personalized recommendation,Matrix factorization,Data sparsity,Tags,Temporal factor

论文评审过程：Received 18 May 2018, Revised 2 November 2018, Accepted 3 November 2018, Available online 5 November 2018, Version of Record 10 November 2018.

论文官网地址：https://doi.org/10.1016/j.eswa.2018.11.003