Individual Diversity Preference Aware Neural Collaborative Filtering

作者：

Highlights：

• IDP-NCF is an in-processing diversified recommendation method.

• Two elements of IDP-NCF are based on a pre-training task on item knowledge graph.

• Experiments indicate that IDP-NCF has achieved both high accuracy and diversity.

• Further experiment reveals that IDP-NCF could recommend well for niche items.

摘要

•IDP-NCF is an in-processing diversified recommendation method.•Two elements of IDP-NCF are based on a pre-training task on item knowledge graph.•Experiments indicate that IDP-NCF has achieved both high accuracy and diversity.•Further experiment reveals that IDP-NCF could recommend well for niche items.

论文关键词：Diversified recommendation,Accuracy,Diversity,Collaborative filtering,Knowledge graph

论文评审过程：Received 25 May 2022, Revised 15 August 2022, Accepted 16 August 2022, Available online 30 August 2022, Version of Record 18 October 2022.

论文官网地址：https://doi.org/10.1016/j.knosys.2022.109730