Towards an intelligent desk matching behaviors and performance of learners

This paper proposes the design of a system to semi-automatically identify possible relationships between learner behavior and task performance in an intelligent desk. The aim of detecting such relationships is to help teachers and students in the learning process, supporting their activities with an unobtrusive observation to identify special needs to improve their performance. The proposed system is based on models of the desk, of behaviors arising from user-object interactions, and task performance. Its components are designed to address four main functions: data acquisition, behavior identification, task performance identification, and computing relationships between behaviors and task performance.