Feature Extraction and Learning Effect Analysis for MOOCS Users Based on Data Mining

Abstract

Feature extraction is an important technology of data mining, it had been widely used in machine learning and pattern recognition. With the increasing number of enrolments in MOOCs, there was a large amount of learning behaviour data generated in MOOCs platforms. Through the data analytics of these learning behaviour data, some useful predictions and guidance information will be obtained. In this paper, two machine learning algorithms based on learner behavioural data is presented. According to pre-extracted features, our algorithms would take the history data into account and can detect changes in learner behaviour over time. In our experiments, we use our proposed algorithms to predict dropout rates of MOOCs. Compared with other existing algorithms in common use, our proposed algorithm can predict dropout accurately better.