Association Rules Mining Method of Big Data for E-Learning Recommendation Engine

Abstract

Today, recommender systems are increasingly used due to its success in several areas such as e-commerce, tourism, social networks, and e-learning. Indeed, most of the computing environment for human learning, especially the online learning platforms have a very large number of learners’ profiles, thousands of courses, and various educational resources. However, students often face many challenges, such as the absence of a real solution of recommendation to orientate them to take more appropriate learning materials. In this article, we develop a recommendation engine for the e-learning platform in order to help learners to easily find the most proper pedagogical resources without any search effort. It aims to discover relationships between student’s courses activities through the association rules mining method. We also focus on the analysis of past historical data of the courses enrollments or log data. The article discusses particularly the frequent itemsets concept to determine the interesting rules among objects in the transaction database. Then, we use the extracted rules to find the list of suitable courses according to the learner’s behaviors and preferences. Next, we implement our system using Apriori algorithm and R, which is efficient big data analysis language and environment, on data collected from ESTenligne [ESTenLigne project is supported by the EST Network of Morocco and the Eomed association (http://www.eomed.org)] platform database of High school of Technology of Fez. Finally, the experimental results prove the effectiveness and reliability of the proposed system to increase the quality of student’s decision, guide them during the learning process and provide targeted online learning courses to meet the needs of the learners.