Analysing Computer Science Course Using Learning Analytics Techniques

The academic community has identified some critical issues facing the sector and are trying to Figure out how these can be addressed. One of these issues is that Science, Technology, Engineering, and Mathematics (STEM) fields have notoriously low retention rates. According to 2016 report on Australia’s STEM workforce only 32% are university qualified. LA has become an emerging research field where knowledge is extracted and patterns are discovered from e-learning systems. LA provides a set of techniques which can help improve the retention rate by identifying students’ needs, and analysing students’ learning behaviour patterns. The objective of our research is to introduce a data-driven investigation to determine students’ learning behaviour in a computer science course. Our research uses a variety of techniques such as Clustering (K-Means), and Classification (Decision Tree Induction) in order to achieve the goal to discover data-driven knowledge from the Moodle Learning Management System (LMS) and the Student Information System (SIS). The analysis was carried out using log data obtained from Moodle and student’s information obtained from the SIS in a computer science course. The study collected demographic profiles of students studying a course in Information Systems and Analysis and compared them to find answers to questions such as how student’s engagement to online activities and accessing learning resources impact students’ success rate and hence improving retention rate by identifying students need on time. For this work, Rapid Miner data mining tool was used to mine and analyse data from the SIS and the Moodle system.