A Predictive Model for Video Lectures Classification

Resumo

In the educational context, it is important to provide students with learning resources, such as tutorials, video lectures, and educational games to help their learning process, especially when they are at home and have difficulties or doubts. In these cases, recommendation systems have been used to suggest learning resources for students, avoiding the difficult task of making the manual process of searching and selecting resources. Most generic recommendation systems for video lectures use viewing history of the user to make recommendations of videos, which that are consistent with the interests of users. In the educational context, other factors must be considered, the video should not only be of interest to the student, as it will be used as a learning resource for the main purpose of helping the student to learn a particular subject or clarify doubts. Hence, in this paper we evaluated three classifiers and propose a predictive model to classify video lectures according to their quality. We applied machine learning algorithms on a set of video lectures by students classified according to some quality requirements. We conducted an experiment and preliminary results indicate good quality of the selected prediction model.