Ecole Polytechnique Federale de Lausanne (EPFL) is a leading engineering school in the world, and one of the two prestigious Swiss Federal Institutes of Technology. The Human Computer Interaction Group is part of EPFL. Its research interests are multi-disciplinary, combining human computer interaction, artificial intelligence, and behavior sciences, with key contributions to recommender technology, behavior analytics, and emotion detection.
Our main contribution to this project is our work on behavior profiling methods and lifestyle recommender systems based on wearable sensor data. We are tackling some of the fundamental issues in analyzing time series data for physical activities: how to model user behaviors to understand the temporal patterns and trends; how to analyze behavior data to identify determinants that set apart users who respond positively to interventions vs. those who do poorly, and how to exploit user differences as well as similarities to recommend achievable behavior changes over time.