Modelling socially-influenced conditional preferences over feature values in recommender systems based on factorised collaborative filtering
作者：
Highlights：
• Novel method for modelling conditional preferences over feature values is proposed.
• The social influence on the components of preferences is taken into consideration.
• Extensive experiments and statistical analysis carried out to evaluate our model.
• Our method achieves statistically significant improvements over recent methods.
摘要
•Novel method for modelling conditional preferences over feature values is proposed.•The social influence on the components of preferences is taken into consideration.•Extensive experiments and statistical analysis carried out to evaluate our model.•Our method achieves statistically significant improvements over recent methods.
论文关键词：Recommender systems,Latent factor models,Probabilistic matrix factorisation,Feature value preferences,Homophily,Social influence
论文评审过程：Received 28 December 2016, Revised 23 May 2017, Accepted 24 May 2017, Available online 7 June 2017, Version of Record 17 July 2017.
论文官网地址：https://doi.org/10.1016/j.eswa.2017.05.058
