×î×¼µÄÁùºÏ²ÊÂÛ̳

Why this research is needed

A vast amount of behavioural and social science research exists. For example over 100 behaviour change intervention evaluation reports published every week. It is not feasible to extract and synthesise all the key information from these reports by hand.

Terms are also used inconsistently and ambiguously in reports describing behaviour change interventions.

Both of these challenges severely limitÌýour ability to build and test models of behaviour. In turn this decreases our ability to predict intervention outcomes in novel scenarios.

Creating an automated knowledge system

A big milestone was releasing the ground-breaking Behaviour Change Intervention Ontology (BCIO). The BCIO provides a knowledge structure to organise and understand evidence about behaviour change interventions.

Machine Learning algorithms, drawing on the BCIO, are creating a Knowledge System. This will automatically identify new, relevant research to predict a range of behaviour change outcomes.

Real world applications

The first use case we developed was smoking cessation. The second, currently underway, is interventions to increase physical activity. Ìý

We have developed a user interface, focused on the smoking cessation use case. This will enable policy-makers, planners, practitioners and others to understand the following question.

When it comes to behaviour change interventions:

• What works,
• compared with what,
• for what behaviours,
• how well,
• for how long,
• with whom,
• in what setting,
• and why?

Outputs

We have created online tools to facilitate research, synthesise evidence and predict outcomes. Resources from the project are on the and . We continue to publish to ourÌý.Ìý

Ìý