Mining Learning Behavioral Patterns of Students by Sequence Analysis in Cloud Classroom

Learning Analytics in MOOCs

Learning analytics has gained increased attention in recent years with the widely use of MOOC (Yousef, Chatti, Schroeder & Wosnitza, 2015). The Society for Learning Analytics Research (SoLAR) defines that “learning analytics is the measurement, collection, analysis and reporting of data about learners and their contexts, for purposes of understanding and optimizing learning and the environments in which it occurs” (LAK2011, 2010).

Some researchers have done a series of studies about the learning analytics in MOOCs during the past two years. For example, Anderson et al. (2014) developed a conceptual framework for understanding how users engaged in MOOCs and examined the different behavioral patterns of high- and low-achieving students. They also investigated how forum participation was related to other modules of the course. They found that students with high grades watched more videos than those with lower grades. Moreover, students with different grades behaved differently in assignment and forum. Kizilcec et al. (2013) presented a classification method that identified a small number of longitudinal engagement trajectories in MOOCs. First, they made a rough description of each student's individual engagement in a course. For each assessment period, all participants were labeled either “on track” (did the assessment on time), “behind” (turned in the assessment late), “auditing” (didn't do the assessment but engaged by watching a video or doing a quiz), or “out” (didn't participate in the course at all). Then they visualized labels sequence to show movement if students moved from one assignment period to next. In these papers, learners are classified based on their patterns of interaction with video lectures and assessments, which is also the primary feature of most MOOCs to date. Halawa et al. (2014) presented a dropout predictor that used student activity features to predict which students were at risk high of dropout. These students usually skipped videos and assignments, rarely logined in the platform for two consecutive weeks, and got a student's average quiz score below 50 points. Wen et al. (2014) from Carnegie Mellon University proposed a novel method to characterize types of sessions in MOOCs by mining the habitual behaviors of students within individual sessions. They modeled learning sessions as a distribution of activities and activity sequences with a topical N-gram model. To summarize the related researches, the mining of learning behavior patterns is a significant procedure of the learning analytics in MOOCs. However, compared with some studies on behavioral patterns, there are few studies exploring students’ behavioral patterns by adopting the method of sequence analysis in MOOCs. In addition, Clould Classroom is a small private online learning platform, which is designed for every grade level’s students at college. It can be considered as a powerful tool for researchers to mine students’ differences in learning behaviors. This paper, therefore, is to identify learning behavior sequence patterns of students for each grade level in a MOOC-style online course platform.