Evaluation of clustering and topic modeling methods over health-related tweets and emails

作者：

Highlights：

• Evaluation of topic modeling and clustering on health-related tweets and emails.

• Topic modeling: LSI, LDA, BTM, GibbsLDA, Online LDA, Online Twitter LDA, and GSDMM.

• Clustering: k -means with two feature representations (TF-IDF and Doc2Vec).

• The evaluation is based on two internal and five external cluster validity indices.

摘要

•Evaluation of topic modeling and clustering on health-related tweets and emails.•Topic modeling: LSI, LDA, BTM, GibbsLDA, Online LDA, Online Twitter LDA, and GSDMM.•Clustering: k -means with two feature representations (TF-IDF and Doc2Vec).•The evaluation is based on two internal and five external cluster validity indices.

论文关键词：Topic modeling,Clustering,Internal cluster indices,External cluster indices,Natural language processing

论文评审过程：Received 31 May 2020, Revised 30 March 2021, Accepted 5 May 2021, Available online 7 May 2021, Version of Record 21 May 2021.

论文官网地址：https://doi.org/10.1016/j.artmed.2021.102096