Suggest what to tag: Recommending more precise hashtags based on users’ dynamic interests and streaming tweet content

作者：

Highlights：

• Design online Twitter-User LDA to capture Twitter users’ dynamic interests.

• Introduce incremental biterm topic model to discover topic distribution of streaming tweet content.

• Combine tweet content and dynamic user interest to build a personalized hashtag recommendation method: User-IBTM.

• To the best of our knowledge, the proposed User-IBTM is the first method which uses online algorithms and topic models for short texts to recommend hashtags in Twitter.

摘要

•Design online Twitter-User LDA to capture Twitter users’ dynamic interests.•Introduce incremental biterm topic model to discover topic distribution of streaming tweet content.•Combine tweet content and dynamic user interest to build a personalized hashtag recommendation method: User-IBTM.•To the best of our knowledge, the proposed User-IBTM is the first method which uses online algorithms and topic models for short texts to recommend hashtags in Twitter.

论文关键词：Recommender systems,Social networks,Micro-blogging,Hashtag recommendation,Topic models,Online algorithms

论文评审过程：Received 11 November 2015, Revised 21 May 2016, Accepted 23 May 2016, Available online 24 May 2016, Version of Record 18 June 2016.

论文官网地址：https://doi.org/10.1016/j.knosys.2016.05.047