Ji, Shaoxiong, et al. "Detecting suicidal ideation with data protection in online communities." International conference on database systems for advanced applications. Cham: Springer International Publishing, 2019.

This study introduces a groundbreaking data-protecting solution aimed at detecting suicidal ideation through virtual intelligent agents, without compromising user privacy. Traditional methods, which aggregate vast amounts of personal data, raise significant privacy concerns. The proposed solution trains a local data-preserving model on each user's device, sharing only model parameters, not personal data, with a central server. This approach utilizes an innovative average difference descent algorithm for optimizing parameter aggregation from various client models. Experimental validation with real-world social community datasets indicates the potential of this technology to revolutionize privacy-conscious mental health services.