Using Machine Learning to Verify the Identity of Distance-Learning Students
Verifying students’ identity in online/distance-learning courses is becoming a serious concern for university professors and administrators. Although many institutions mandate that final exams must be written in person at the university campus or under the supervision of a proctor for online courses, many universities sill rely on student login credentials only for submitting course assignments. With the growing demand and popularity of online learning degrees, an efficient students’ authentication system has become vital to maintain the academic integrity of online education programs and the reputation of the institutional credentials.
One approach to address this issue is using behavioural biometrics to verify students’ identity. More specifically, verifying students’ identity by analyzing the manner and the rhythm of their typing on a keyboard, which is knows as keystroke dynamics. That means, in addition to providing the correct password, students’ typing patterns must match their profiles. We developed a machine learning solution to authenticate students using keystroke dynamics. In this session, I’ll present our research work that was conducted to improve the accuracy of Machine Learning classifiers and will demonstrate our smartphone app that authenticates users based on their behavioural biometrics. The audience will be invited to participate in the live demo.
Haytham El Miligi
Assistant Professor, Thompson Rivers University