A learning styles based recommender system prototype for edX courses

Abstract

Recommender Systems (RSs) are information filtering systems that gather data about users, to build useful recommendations for them. In todays age of bigdata, RSs have become important tools for finding relevant items from numerous options. RSs for web-based learning assist learners in discovering relevant learning content, according to requirements. Since learners have diverse needs and contexts, recommendations need to be highly personalized. In web-based learning, MOOCs are globally accessible online courses, designed to support learning of a large number of students of different backgrounds. For modelling users, existing MOOC RSs focus on extrinsic properties such as demographics, ratings, whereas a students learning is more dependent on intrinsic qualities like competency, learning styles. In this work, a new RS prototype has been designed, for recommending MOOCs based on requirements and learning styles. This prototype is based on an adapted version of the FSLSM LS model and works exclusively for edX MOOCs.