Tracking user-role evolution via topic modeling in community question answering

作者：

Highlights：

• We have tracked user-role evolution and investigated its influences on the performance of question recommendation in CQA.

• We proposed a TRM to track user-role evolution. With different independence assumptions, we developed two variants: the independent time-aware role model (ITRM) and dependent time-aware role model (DTRM).

• We carried out experiments on real-world data from Stack Overflow. The results show that TRM consistently and significantly improves the performance of question recommendation.

摘要

•We have tracked user-role evolution and investigated its influences on the performance of question recommendation in CQA.•We proposed a TRM to track user-role evolution. With different independence assumptions, we developed two variants: the independent time-aware role model (ITRM) and dependent time-aware role model (DTRM).•We carried out experiments on real-world data from Stack Overflow. The results show that TRM consistently and significantly improves the performance of question recommendation.

论文关键词：Community question answering,Role evolution,Time-aware role model

论文评审过程：Received 26 August 2018, Revised 26 May 2019, Accepted 2 July 2019, Available online 13 July 2019, Version of Record 13 July 2019.

论文官网地址：https://doi.org/10.1016/j.ipm.2019.102075