Recipient organisationUniversity of ExeterSource-published name: University of Exeter
Funding£97K
PeriodFeb 2025 — Feb 2028
In plain English
AI plain-English summary
Only one in four people with depression worldwide receives proper care, and this project builds AI tools to reach the other three. Depression is the second-largest contributor to lost healthy life years globally, yet most sufferers never get treatment. The project tackles two stubborn technical gaps that block AI from helping: first, people describing depression often use figurative language—metaphors, irony, indirect expressions—which current language models handle poorly; second, effective support must be personalised to each person’s specific life circumstances, not generic. The researchers will develop natural language processing systems that can interpret abstract self-narratives and generate tailored, real-time interventions. If successful, the system could provide mental health support at scale to populations that currently have no access—particularly in Brazil and the UK, where social determinants such as poverty and housing are known risk factors. The work also aims to clarify how those social factors shape mental health in two very different national contexts, informing public policy decisions. The immediate output is a set of computational methods, not a deployed app, but the fundamental science here—teaching machines to understand how people actually talk about suffering—could underpin future autonomous, ethical, and engaging mental health tools.
View original technical description
Depression is a mental health disorder that affects a large portion of the global population, being the second largest contributor to a decrease in healthy life expectancy. Depression is characterised by a clinically significant form of psychological suffering that leads to significant impairment in someone's functionality, reduced quality of life and, in severe cases, can lead to death due to the risk of suicide. However, according to the World Health Organization, only a quarter of individuals suffering from mental health disorders receive proper care. Advances in Artificial Intelligence (AI) and Natural Language Processing (NLP) research have been developed to a level that can be used for proposing computational solutions that assist in the detection and intervention in mental health conditions. AI and NLP based solutions that aid in the identification of signs of depression can be useful both in individual treatment and in making public policy decisions. Similarly, solutions that offer autonomous, ethical, reliable, controlled, and engaging intervention, in real time, can help mitigate the damage caused by depression. This project works on proposing and developing AI and NLP based solutions for the detection and intervention of mental health conditions that can have a broader reach and allow mental health support to individuals and populations that would not otherwise have access to it. Furthermore, as social determinants are frequently mentioned as risk factors for mental health conditions, this project also aims at furthering the understanding about them in two contexts (Brazil and the United Kingdom). This project aims to address scientific challenges that are still present and very relevant in this context: (i) dealing with more abstract language (such as figurative language) commonly used in mental health self-narratives, and (ii) outputting personalised interventions suitable for an individual's context.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know