Explainable link prediction based on multi-granularity relation-embedded representation

作者：

Highlights：

• A social relation extraction method is proposed via semantic association of texts.

• Multi-granular semantic specificity embeddings are used to predict social relation.

• Social relation is explained from the aspects of multi-granularity interest.

摘要

•A social relation extraction method is proposed via semantic association of texts.•Multi-granular semantic specificity embeddings are used to predict social relation.•Social relation is explained from the aspects of multi-granularity interest.

论文关键词：Social relation extraction,User interest profile,Multi-granularity relation-embedded representation,Explanation

论文评审过程：Received 3 June 2020, Revised 14 April 2021, Accepted 14 August 2021, Available online 17 August 2021, Version of Record 26 August 2021.

论文官网地址：https://doi.org/10.1016/j.knosys.2021.107402