Using Association Rule Mining Algorithm to Improve the Order of Content Delivery in CS1 Course

• Published in: 2019 IEEE Frontiers in Education Conference (FIE) pp. 1 - 5
• Main Authors: Kaur, Rupinder, Brown, Tamaike, Walia, Gursimran, Singh, Maninder, Reddy, Mourya
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01-10-2019
• Subjects: Arrays; Association rule mining; computer education; Education; introductory programming; Java; Libraries; machine learning; Programming profession
• ISSN: 2377-634X
• DOI: 10.1109/FIE43999.2019.9028452

This work in progress research paper discusses importance of appropriate order of content delivery in an introductory programming course (CS1). A majority of students face problems when programming concepts are introduced to them in disorderly fashion, reducing their retention. It has been observed that the instructors use their intrinsic judgement to order course contents without considering effective student learning outcome. This traditional approach of teaching a course in an unstructured way leads to high failure and dropout rate in CS1 course. In this study, an association rule mining (ARM) based approach is being used to understand the order of information that students should be exposed to in CS1 courses. Our proposed approach follows an empirical research methodology and generates strong ARM rules. These rules can assist instructors to structure CS1 topics and reflect on required prerequisites for problematic topics to improve students' learning. Our initial results and observations yielded promising and logical inferences and uncovered few anomalies with traditional pedagogy during course offering.