A mental state Knowledge–aware and Contrastive Network for early stress and depression detection on social media

作者:

Highlights:

• Mental state commonsense knowledge from COMET to model the mental states of speakers.

• Knowledge-aware mentalisation to enhance the ability of understanding mental states.

• Supervised contrastive learning to capture class-specific semantic features.

• SOTA on a depression detection, stress detection, stress factors recognition dataset.

摘要

•Mental state commonsense knowledge from COMET to model the mental states of speakers.•Knowledge-aware mentalisation to enhance the ability of understanding mental states.•Supervised contrastive learning to capture class-specific semantic features.•SOTA on a depression detection, stress detection, stress factors recognition dataset.

论文关键词:Mental health,Natural language processing,Mental state knowledge,Mentalisation,Contrastive learning

论文评审过程:Received 5 March 2022, Revised 19 April 2022, Accepted 24 April 2022, Available online 6 May 2022, Version of Record 6 May 2022.

论文官网地址:https://doi.org/10.1016/j.ipm.2022.102961