Analyzing Students' Interaction Based on Their Response to Determine Their Learning Outcomes

Online learning platforms such as Moodle and MOOC (Coursera, edX, etc) have become popular in higher education. These platforms provide information that are potentially useful in developing new student learning models. One source of information provided by these platforms is in the form of student interaction with one another, instructors, and the platform itself. These interactions contain various activities such as: participation in forum discussion, how frequently a student is logged into their account, and frequency of reading posted activities, etc. Using Data Mining techniques, namely clustering algorithms to find students with similar behavior patterns, our goal is to develop a student model that can be conducted by learning these interaction patterns. In doing so, we aim to develop a method by which to provide students with different guidelines and instructions that will help to improve their performance. This research is in progress and our data include Moodle online courses in computer science in different semesters.