Graph neural news recommendation with long-term and short-term interest modeling

作者：

Highlights：

• We propose a novel graph neural news recommendation model GNewsRec with long-term and short-term user interest modeling.

• Our model constructs a heterogeneous user-news-topic graph and then applies graph convolution networks to learn user and news embeddings with high-order information endcoded by propagating embeddings over the graph.

• Our proposed model significantly outperforms state-of-the-art method on news recommendation.

摘要

•We propose a novel graph neural news recommendation model GNewsRec with long-term and short-term user interest modeling.•Our model constructs a heterogeneous user-news-topic graph and then applies graph convolution networks to learn user and news embeddings with high-order information endcoded by propagating embeddings over the graph.•Our proposed model significantly outperforms state-of-the-art method on news recommendation.

论文关键词：News recommendation,Graph neural networks,Long-term interest,Short-term interest

论文评审过程：Received 29 July 2019, Revised 20 September 2019, Accepted 7 October 2019, Available online 15 November 2019, Version of Record 15 November 2019.

论文官网地址：https://doi.org/10.1016/j.ipm.2019.102142