Students' Attention Assessment in eLearning based on Machine Learning

Abstract

Multimedia-based Electronic learning (eLearning) is an effective method of knowledge transfer. It provides the opportunity that students can use the videos or other materials at any time after they are delivered. Multimedia applications provide convenience, but there also exist challenges. One of the challenges is to measure and assess students' attention when they are studying online. This paper presents a framework based on machine learning methods for the measurement of students' attention. The framework employs a Gabor wavelet to extract the eye state features and train the model using support vector machines (SVM) to complete automatic classification on students' eye states. Experiments over thousands of facial photos show that the proposed system reaches a good performance, which has a significant value in real applications.