A novel term weighting scheme for text classification: TF-MONO

作者：

Highlights：

• Two novel term weighting schemes are proposed for text classification.

• The proposed schemes are named as TF-MONO and SRTF-MONO.

• The proposed schemes are tested on 3 different benchmark datasets.

• SRTF_MONO method generally outperformed all other schemes.

• TF-MONO has promising results compared to others especially on benchmark datasets.

摘要

•Two novel term weighting schemes are proposed for text classification.•The proposed schemes are named as TF-MONO and SRTF-MONO.•The proposed schemes are tested on 3 different benchmark datasets.•SRTF_MONO method generally outperformed all other schemes.•TF-MONO has promising results compared to others especially on benchmark datasets.

论文关键词：Text classification,Supervised term weighting,Max-occurrence,Non-occurrence

论文评审过程：Received 7 February 2020, Revised 6 July 2020, Accepted 8 July 2020, Available online 24 July 2020, Version of Record 24 July 2020.

论文官网地址：https://doi.org/10.1016/j.joi.2020.101076