DEPTWEET: A typology for social media texts to detect depression severities
作者：
Highlights：
• Tweets with depression symptoms were extracted based on clinical assessment tools.
• Collected tweets were annotated by trained annotators, supervised by domain experts.
• Four depression labels: None, Mild, Moderate, and Severe were added to each tweet.
• A dataset of 40191 tweets on severities of depression is made publicly available.
• The utility of the dataset was validated using statistical and mathematical modeling.
摘要
•Tweets with depression symptoms were extracted based on clinical assessment tools.•Collected tweets were annotated by trained annotators, supervised by domain experts.•Four depression labels: None, Mild, Moderate, and Severe were added to each tweet.•A dataset of 40191 tweets on severities of depression is made publicly available.•The utility of the dataset was validated using statistical and mathematical modeling.
论文关键词：Social media,Mental health,Depression severity,Dataset
论文评审过程：Received 15 February 2022, Revised 8 August 2022, Accepted 25 September 2022, Available online 28 September 2022, Version of Record 7 October 2022.
论文官网地址：https://doi.org/10.1016/j.chb.2022.107503
