Research on MOOC Learners’ Behavior in Discussion Area Based on Text Mining Technology
Abstract
In recent years, MOOC, as a kind of online learning, is becoming increasingly popular among students. Learners communicate mainly through the discussion area while learning in MOOCs. Therefore, MOOC forums play an important role in learning. In this paper, the text mining is adopted to search for the speech data in the discussion area of 39 online learning course, the segmentation algorithm is used to study the emotion, frequency and quality of the speech, furthermore, K means clustering algorithm is designed to classify users by their performance. The results show that users with different identities have a wide range of roles in the discussion forum. Therefore, some users have higher quality of speeches, but the overall speeches is of poor quality. There is also a big gap between the numbers of speeches in various courses.
Keywords
Learners’ behavior, MOOC discussion forum, Text mining, Hidden Markov segmentation model, Cosine similarity of high dimensional space, K-Means Clustering
DOI
10.12783/dtssehs/icems2018/20108
10.12783/dtssehs/icems2018/20108