Health and Wellbeing

Understanding how people’s behaviour changes over time is crucial, particularly in monitoring mental health conditions.

We are creating new methods to track the progress of an individual over time, based on their language use and other interactions on digital media. We work with data from social media, counselling sessions and conversations with carers. Our models can capture sudden changes in mood (switches) as well as gradual changes (escalations). We have shown how capturing such changes over time can help predict risk factors like suicidality.

We are working on better personalisation methods and the link between clinical events (e.g. reported self-harm), changes in mood and changes in symptoms and behaviour. We are developing novel methods for summarising these changes in the broader context of people’s lives. We have used AI techniques to identify the changes in language use characteristic of language disorders. The data we work with is sensitive and we are developing methods for synthetic language data generation and privacy preservation, both to improve the available datasets and to protect the identity of individuals. We envisage that our work will boost AI uptake within mental health and significantly improve mental health monitoring.