Ground-breaking Artificial Intelligence system for behaviour change

Ground-breaking Artificial Intelligence system for behaviour change

Researchers from University College London, the University of Aberdeen, the University of Cambridge and IBM Research have released a collection of five papers from the Human Behaviour-Change Project (HBCP) in Wellcome Open Research.

HBCP is a Wellcome-funded collaborative project to support the application of behaviour change interventions in policy and practice using cutting-edge Artificial Intelligence (AI). Led by Professor Susan Michie, Director of the UCL Centre for Behaviour Change, the project will automate the collation, synthesising and interpretation of a vast and growing body of evidence about behaviour change interventions in order to predict new scenarios to answer users’ questions.   

Behaviour change is vital to solving many of the challenges facing human societies, such as reducing carbon emissions, preventing overuse of antibiotics, stopping the smoking and obesity epidemics and reducing transmission of infectious diseases. A huge amount of information is gathered on how best to do this in different situations but, because there is so much of it, researchers, policy-makers and practitioners have very limited capacity to collate, synthesise and use it.    

The project is a collaboration between behavioural and computer scientists and system architects.  It will create an AI-based Knowledge System that will scan the world’s published reports of behavioural intervention evaluations. This system will extract and analyse relevant information on interventions and their effectiveness using a ‘Behaviour Change Intervention Ontology’, developed as part of the project.

Users will be able to ask questions of the Knowledge System and will receive answers about what interventions are likely to work in given situations.  It will also tell them the level of confidence in its answers and explain the process behind its answers. The first topic being tackled is about smoking cessation. 

The parts of the Behaviour Change Intervention Ontology that have been completed and are published as the first papers in a collection within Wellcome Open Research.  This is a platform where all articles are made publicly available upon submission, before transparent peer review and the publication of a final Open Access version. The initial launch of the HBCP collection contains five papers: 

  • Introduction to the project. 
  • Methodology - explaining the methods we used for ontology development. 
  • Upper-level Ontology - specifying the overarching structure of the Behaviour Change Intervention Ontology (BCIO). 
  • Mode of Delivery Ontology – describing a part of the BCIO that characterises ways that behaviour change interventions are delivered (e.g. by face-to-face contact, websites, video). 
  • Setting Ontology - describing a part of the BCIO that characterises the locations in which interventions are delivered (e.g. what country they are in, whether they are in hospitals or primary care). 

Papers will continue to be published in the collection as other parts of the project are completed, with several currently in the pipeline. More information on the Human Behaviour-Change Project can be found at www.humanbehaviourchange.org or on Twitter @HBCProject