Collaborative filtering recommendation algorithm combining time weight and trust relationship

• Published in: Ji suan ji ying yong yan jiu Vol. 32; no. 12; pp. 3565 - 3568
• Main Authors: Zhao, Haiyan, Hou, Jingde, Chen, Qingkui
• Format: Journal Article
• Language: Chinese
• Published: 01.12.2015
• Subjects: Algorithms; End users; Filtering; Filtration; Mathematical analysis; Recommendations; Recommender systems; Similarity
• ISSN: 1001-3695
• DOI: 10.3969/j.issn.1001-3695.2015.12.008

For the reason that traditional collaborative filtering algorithms only consider the single user rating matrix in calculating users' similarity, this paper proposed a collaborative filtering recommendation algorithm combining time weight and trust relationship TTCF, which first used tags' popularity to portray users' preferences for resources, and used users' time behavior information to obtain users' interest similarity, then considered the trust relationship between users, and used the primary and secondary friends to extend users' familiarity similarity and then got the end users' similarity with interest similarity and familiarity similarity, and at last, combining with users' similarity and time information to generate recommendations for users. The experimental results on the dataset of Last. fm show that the algorithm has a better recommendation results.