A novel community answer matching approach based on phrase fusion heterogeneous information network

作者：

Highlights：

• We define the distance of entities with the same or different types in HIN and propose a novel Type-constrained Top-k similarity entity finding algorithm (TTSEF) for answer selection, which innovatively combines entity attributes and semantic features to achieve answer selection in CQA.

• Abundant experimental demonstrate that proposed algorithm precedes the state-of-the-art similar entity matching methods in CQA.

• A meta-path analysis of the optimal matching answers proves that phrase can serve as a bridge to connect different types of entities in CQA effectively.

摘要

•To the best of our knowledge, it is the first work to propose the phrase information network and employ it to construct a fusion heterogeneous information network (HIN) to represent complex entity relationships in community question answering (CQA).•We define the distance of entities with the same or different types in HIN and propose a novel Type-constrained Top-k similarity entity finding algorithm (TTSEF) for answer selection, which innovatively combines entity attributes and semantic features to achieve answer selection in CQA.•Abundant experimental demonstrate that proposed algorithm precedes the state-of-the-art similar entity matching methods in CQA.•A meta-path analysis of the optimal matching answers proves that phrase can serve as a bridge to connect different types of entities in CQA effectively.

论文关键词：Community question answering,Heterogeneous information network fusion,Phrase embedding,Related entity matching

论文评审过程：Received 2 April 2020, Revised 31 August 2020, Accepted 30 September 2020, Available online 17 October 2020, Version of Record 17 October 2020.

论文官网地址：https://doi.org/10.1016/j.ipm.2020.102408