Graph-based collaborative ranking
作者：
Highlights：
• GRank is a novel framework, designed for recommendation based on rank data.
• GRank handles the sparsity problem of neighbor-based collaborative ranking.
• GRank uses the novel TPG graph structure to model users’ choice context.
• GRank directly ranks items for a target user using personalized PageRank in TPG.
• GRank improves NDCG@10 up to 9% compared to other collaborative ranking methods.
摘要
•GRank is a novel framework, designed for recommendation based on rank data.•GRank handles the sparsity problem of neighbor-based collaborative ranking.•GRank uses the novel TPG graph structure to model users’ choice context.•GRank directly ranks items for a target user using personalized PageRank in TPG.•GRank improves NDCG@10 up to 9% compared to other collaborative ranking methods.
论文关键词：Collaborative ranking,Pairwise preferences,Graph modelling,Recommendation systems,Personalized PageRank
论文评审过程：Received 16 March 2016, Revised 7 September 2016, Accepted 8 September 2016, Available online 9 September 2016, Version of Record 24 September 2016.
论文官网地址：https://doi.org/10.1016/j.eswa.2016.09.013
