ABSTRACT
This study was conducted with data mining (DM) techniques to analyze various patterns
of online learning behaviors, and to make predictions on learning outcomes. Statistical
models and machine learning DM techniques were conducted to analyze 17,934 server
logs to investigate 98 undergraduate students’ learning behaviors in an online business
course in Taiwan. The study scientifically identified students’ behavioral patterns and
preferences in the online learning processes, differentiated active and passive learners,
and found important parameters for performance prediction. The results also
demonstrated how data mining techniques might be utilized to help improve online
teaching and learning with suggestions for online instructors, instructional designers and
courseware developers.
Keywords: Educational data mining, online learning, online teaching, online learning
behavioral patterns, online learning activity patterns, online learning predictions,
prediction model, clustering analysis, association analysis, decision tree analysis