A study of supervised term weighting scheme for sentiment analysis

作者：

Highlights：

• We propose a supervised term weighting scheme for sentiment analysis.

• This scheme use ITD and ITS to measure the importance of a term in sentiment analysis.

• Experiment results on three real datasets show that our scheme outperforms state-of-the-art unsupervised approaches.

摘要

•We propose a supervised term weighting scheme for sentiment analysis.•This scheme use ITD and ITS to measure the importance of a term in sentiment analysis.•Experiment results on three real datasets show that our scheme outperforms state-of-the-art unsupervised approaches.

论文关键词：Sentiment analysis,Term weighting,Supervised learning,Experimentation,Performance

论文评审过程：Available online 5 November 2013.

论文官网地址：https://doi.org/10.1016/j.eswa.2013.10.056