Artificial Intelligence and Natural Language Processing: New Frontiers in the Detection and Monitoring of Suicidal Behavior in Adolescents

Suicidal behavior in adolescence represents a critical public health challenge, aggravated by barriers to early detection and access to specialized services. This symposium presents cutting-edge advances at the intersection of Clinical Psychology and Artificial Intelligence, focusing on the use of Natural Language Processing (NLP) models to optimize the identification and monitoring of suicide risk in young populations.

Through four complementary presentations, the symposium explores the full cycle of technological intervention. From the methodological and ethical framework of large multidisciplinary projects, Dr. De la Torre will present evidence on the assessment of discourse-based markers derived from natural language in adolescents at risk of emotional disorders and suicidal behavior (ALENTAR-J-CM Project). Additionally, Adriana García will present a study on the use of machine learning techniques to understand complex phenomena such as non-attendance at psychological evaluations, analyzing the impact of family dynamics. Dr. Doval will present data on specific applications of transformer-based models (such as BERT) to improve clinical accuracy in hospital emergency settings involving suicide attempts. Finally, Dr. Scalingi will address the integration of clinical and linguistic variables to predict suicidal behavior among adolescents hospitalized following a suicide attempt.