Improving personalized recommendations using community membership information
作者：
Highlights：
• This study explores community membership as a useful information source.
• The first part examines the presence of shared interests among community co-members.
• The second part examines recommendations based on users’ community membership.
• Recommendations based on community membership are efficient and complete.
• For cold-start users, recommendations using community membership are highly valuable.
摘要
•This study explores community membership as a useful information source.•The first part examines the presence of shared interests among community co-members.•The second part examines recommendations based on users’ community membership.•Recommendations based on community membership are efficient and complete.•For cold-start users, recommendations using community membership are highly valuable.
论文关键词：Social network-based recommendations,Personalized recommendations,Online community,Online group,CiteULike
论文评审过程：Received 27 October 2016, Revised 20 May 2017, Accepted 28 May 2017, Available online 5 June 2017, Version of Record 5 June 2017.
论文官网地址：https://doi.org/10.1016/j.ipm.2017.05.005
