The small research group leaded by Romain Zeiliger, within the GATE-CNRS lab focuses on innovative software design: advanced browsers, intelligent agents, new user interfaces, etc.
These innovative tools are always designed according to a social and cognitive approach.
Iwas involved in the ACCEL project, an intelligent agent for accelerating user interaction. It consists in learning user behavior in order to automatically detect repetitive sequence and propose the user to complete the task (anticipation shortcuts). In particular, I designed the algorithm that allows the agent to detect repetitive tasks in user behavior.
no published paper, sorry!