New Attitude to Learning in Virtual Environments - Mining Physiological Data for Automated Feedback

摘要

We present the great potential of real time data, obtained from 1/ mouse and keyboard events' monitoring, 2/ physiological measurements of the individualized reaction on the learning process of the learning subject in virtual learning environments. We emphasized simple, non invasive, easily available methods like eyes tracking, blink rate and blink speed measurements, elec-trodermal activities measurements, and heart and/or respiration rate. Those methods have big potential to reflect decreasing attention, increasing visual or cognitive information load, task difficulty, tension, stress and fatigue. We compared the 'real time' data with records obtained from screen captivate SW, video and audio records, records of external observers and learners (volunteers) interviews. We highlight the advantages and constraints of different data acquisition approaches, as well as constraints, done by hardware and software limits, and discuss the future potential for automated learners' feedback within VLE.