Italiano (IT) 
English (UK) 
We investigate how to effectively promote physical activity with message-based interventions, combining the explanatory power of theory-driven structural equation modelling with the predictive power of data-driven artificial intelligence.
We test the effectiveness of messages transmitted through the PsyMe app and aimed at promoting physical activity.
The psychological profile of people is traced, to identify the most suitable messages to promote adherence to physical activity programs.
Then, through a Deep Reinforcement Learning (DRL) approach, we achieve an automated, fast-profiling policy to quickly select the best messaging strategy, based on the features of each possible respondent.
Finally, the fast-profiling method is integrated into an AI-based chatbot.
Research group
Mauro Bertolotti, Valentina Carfora, Marco Piastra, Patrizia Catellani
Funding
The project has been funded under the Hub Innovation call by the Lombardy Region, and has been carried out in collaboration with the University of Pavia, the Maugeri Clinical Institutes, and Athics
Media
The role of psychology in the RE-Hub-ILITY Project
Presentation of the results of AI & Physical Activity project during the final meeting of the RE-Hub-ILITY Project – Fondazione Maugeri
RE-Hub-ILITY project presentation
29 November 2022. Presentation of the project RE-Hub-ILITY aiming to implement an integrated system for home rehabilitation using the best existing technologies
Slide
Publications
Applying psychology of persuasion to conversational agents through reinforcement learning: An exploratory study
Di Massimo, F., Carfora, V., Catellani, P., & Piastra, M. (2019). CEUR – Workshop Proceedings, 2481, 27. 
Dialogue management in conversational agents through psychology of persuasion and machine learning
Carfora, V., Di Massimo, F., Rastelli, R., Catellani, P., & Piastra, M. (2020). Multimedia Tools and Applications, 39, 35949-35971.
