Item feature refinement using matrix factorization and boosted learning based user profile generation for content-based recommender systems

作者：

Highlights：

• The model applies item feature refinement before applying content-based filtering.

• The use of matrix factorization avoids the sparsity in item feature information.

• The refined item features are used for generating similarity information.

• User profiles are built using the ensemble learning technique for recommendation.

摘要

•The model applies item feature refinement before applying content-based filtering.•The use of matrix factorization avoids the sparsity in item feature information.•The refined item features are used for generating similarity information.•User profiles are built using the ensemble learning technique for recommendation.

论文关键词：Content-based filtering,Ensemble learning,Matrix factorization

论文评审过程：Received 27 December 2021, Revised 14 May 2022, Accepted 9 June 2022, Available online 17 June 2022, Version of Record 23 June 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.117849