This presentation, prepared by Dr Clare Llewellyn, a lecturer in Governance, Technology and Data at the University of Edinburgh, outlines the methods and key findings of our project with UNICEF which investigated developing a methodology for using AI to identify social media discussions on mental health and well-being from young people and adolescents.  

Building on UNICEF's preexisting work in Ukraine in 2021-22, this project took a more participatory approach with adolescents and young people in Tajikistan and Kazakhstan to better understand online discourse around mental health among adolescents.  

The presentation explains how the outcomes from youth engagement, which focused on understanding what kind of language young people use online, and which platforms they use in the context of mental health and well-being, contributed to the wider work using AI models. Dr Llewellyn, the principal investigator on the project, touches on the team’s innovative use of a Large Language Model (LLM) to identify relevant content from large data sets.  

The presentation discusses the methodology's potential and scalability, as well as its limitations.