Fine-grained depression analysis based on Chinese micro-blog reviews
作者：
Highlights：
• We propose an expert-annotated dataset for fine-grained depression analysis.
• The dataset has practical significance for sociology and health care.
• We propose a novel model for joint depression degree and cause prediction.
• The proposed method outperforms several strong baselines and keeps interpretability.
• We conduct detailed analyses for better understanding the data and model.
摘要
•We propose an expert-annotated dataset for fine-grained depression analysis.•The dataset has practical significance for sociology and health care.•We propose a novel model for joint depression degree and cause prediction.•The proposed method outperforms several strong baselines and keeps interpretability.•We conduct detailed analyses for better understanding the data and model.
论文关键词：Natural language processing,Depression analysis,Social media,Multi-task learning,BERT
论文评审过程：Received 15 January 2021, Revised 10 June 2021, Accepted 29 June 2021, Available online 21 July 2021, Version of Record 21 July 2021.
论文官网地址：https://doi.org/10.1016/j.ipm.2021.102681
