Enhancing Collaborative Filtering of Learning Resources with Semantically-Enhanced Social Tags

With ubiquitous access to Web 2.0 applications, users on social networks can now collaboratively author and share hypermedia learning resources, contributing to more engaging learning experiences in Technology-Enhanced Learning outside formal education. Though these resources are ready to be harnessed for educational purposes, they may or may not be educational or may be suitable for some users only. Hence, it is crucial to annotate these resources to personalize these resources to different user needs. We propose to annotate learning resources with folksonomy-derived description metadata for personalization to context profiles of users. A semantic model for folksonomy was formulated to integrate controlled vocabularies in folksonomy. It not only reduces reliance on domain experts in annotating learning resources but also opens new avenues for more comprehensive resource description, compared to numerical ratings in resource recommendation algorithm. The working principle of the resource annotation and recommendation mechanism is demonstrated via a prototype system implemented in a social network environment. A pilot user study shows that learners are positive about the system in terms of aspects apart from accuracy.