A Conceptual Framework for Detection of Learning Style from Facial Expressions using Convolutional Neural Network

Abstract

There are millions of learning materials over the internet that students can use to assimilate new information. But once their preferred learning style is known, they can be provided with a responsive recommendation so that can focus more on representations that will foster their understanding. Providing students with preferred learning object no doubt increase their motivation and hence their learning outcome. Identifying student’s learning styles allows them to learn better and faster through several means. Traditionally, a test (use of questionnaire) is usually conducted for automatic detection and prediction of student’s learning preferences particularly in e-learning. This approach though valid and reliable in detection of learning styles, but it is also associated with many challenges; learner self-report bias, individual earning styles may vary over time, Students not aware of the importance or the future uses of the questionnaire. To this end, this paper proposed a conceptual framework for detection of learning style from facial expression using Convolution neural network.