A multiview graph collaborative filtering by incorporating homogeneous and heterogeneous signals

作者：

Highlights：

• Homogeneous and heterogeneous signals are leveraged for collaborative embedding.

• The consistency of attribute and neighbor view-specific node embeddings is studied.

• A multiview graph collaborative filtering framework is designed for CTR prediction.

• The validity of data sparse recommendation is verified by extensive experiments.

摘要

•Homogeneous and heterogeneous signals are leveraged for collaborative embedding.•The consistency of attribute and neighbor view-specific node embeddings is studied.•A multiview graph collaborative filtering framework is designed for CTR prediction.•The validity of data sparse recommendation is verified by extensive experiments.

论文关键词：Recommender systems,Attribute embedding,Neighbor embedding,Multiview graph collaborative filtering

论文评审过程：Received 14 April 2022, Revised 30 July 2022, Accepted 22 August 2022, Available online 13 September 2022, Version of Record 13 September 2022.

论文官网地址：https://doi.org/10.1016/j.ipm.2022.103072