Emotion detection in suicide notes

作者：

Highlights：

• Suicide note analysis is a useful tool towards suicide prevention.

• Automatic techniques are crucial for analyzing large amounts of online text.

• Fine-grained emotion detection is possible using lexico-semantic features.

• There is a clear correlation between performance and data availability.

• Detection performance improves when data is preprocessed with spelling correction.

摘要

•Suicide note analysis is a useful tool towards suicide prevention.•Automatic techniques are crucial for analyzing large amounts of online text.•Fine-grained emotion detection is possible using lexico-semantic features.•There is a clear correlation between performance and data availability.•Detection performance improves when data is preprocessed with spelling correction.

论文关键词：Natural language processing,Suicide,Emotion

论文评审过程：Available online 28 May 2013.

论文官网地址：https://doi.org/10.1016/j.eswa.2013.05.050