The group is working on several projects including:

• AI assistance to AI: Many tasks require the collection and analysis of data. However, for a non-specialist user, these can be difficult. We aim to make more interactive AI systems that does not only execute the user's orders, but also try to anticipate them, to understand the user's goals and to contribute to solving the overall task in a more collaborative way. 
• AI for Intelligent Tutoring Systems: Most vocabulary learning applications rely on a heuristic called Leitner system. As an alternative, we develop a model-based system that relies on a cognitive model of how the learner learns, forgets and reasons, in order to better select appropriate learning material for them. 
• Analogical and case-based reasoning: An essential characteristic of humans is their ability to reason by analogy and to adapt their knowledge to new cases. This ability is still difficult to reach with AI systems, though, and we aim to implement this form of reasoning into AI systems. This can be used in particular to better understand humans' actions in assistance scenarios. 

A list of recent publications can be found here.