A MOOC courses recommendation system based on learning behaviours

MOOC1 courses recommendation is an important and challenging task, especially in an era with a quick development of Internet which consists of gigantic and diverse education resources. Its challenge is due to its massive amount of education information in almost all academic fields and as a result, the inevitable negligence of personalized needs for certain knowledge. Therefore, the research on timely capturing of the learners' behaviours and then personalized guidance of their learning process becomes increasingly essential. In this paper, we analyse online learning behaviours to improve personalized recommendations in MOOC courses. Our main contribution is to utilize information from different sources and design a centralized framework to combine them, thus making superior recommendation. We propose two different models based on the above sources and a combined model, and then contrast the models with other traditional models to prove the superior performance of our models.