The computer science team at IBM Research Dublin is working to develop information extraction algorithms to (semi)-automatically extract key pieces of information from behaviour change intervention evaluation reports. This information is populating a central behaviour change intervention database structured according to the Behaviour Change Intervention Ontology (BCIO). Additionally, the team is working to develop Machine Learning and reasoning algorithms to perform inference on the central database to generate recommendations for future interventions, and predictions about the likely outcomes of as yet untested interventions.