Authenticating student work in an e-learning programme via speaker recognition

The past decade has seen the proliferation of elearning and distance learning programs across a wealth of discipline areas. In order to preserve maximum flexibility in outreach, student assessment based exclusively on remotely submitted work has become commonplace. However, there is also growing evidence that e-learning also provides increased opportunity for plagiarism, with obvious consequences for learning effectiveness. This paper reports on the development of a prototype student authentication system, designed for use with a graduate e-learning program. The proposed system can be used to authenticate telephone-based oral examination which can, in turn, be used to confirm a student's ability in relation to submitted assignments and on-line test results. The prototype low-cost system is shown to be sufficiently accurate to act as an effective deterrent against plagiarism.