5ald/whatsapp-depression-synthetic

WhatsApp抑郁症检测数据集是一个用于二进制文本分类任务的数据集，专注于在非正式的WhatsApp和SMS风格文本中检测抑郁症。数据集包含两个子集：一个是由AI模型生成的合成子集，另一个是由数据集作者实际WhatsApp消息组成的真实子集。合成子集进一步分为训练集、验证集和测试集，而真实子集仅用于评估目的。数据集为英文，包含标记为抑郁(1)或非抑郁(0)的消息。README还详细说明了生成和收集过程、预期用途、局限性、伦理考虑和引用信息。

The WhatsApp Depression Detection dataset is designed for binary text classification tasks related to mental health, specifically detecting depression in informal WhatsApp and SMS-style text. It comprises two subsets: a synthetic subset generated by an AI model and a real subset consisting of actual WhatsApp messages from the dataset author. The synthetic subset is split into training, validation, and test sets, while the real subset is reserved for evaluation only. The dataset is in English and includes labels indicating depressive (1) or non-depressive (0) messages. The README details the generation and collection procedures, intended use, limitations, ethical considerations, and citation information.