RECOMMENDER PLATFORM FOR EDUCATIONAL CONTENT THROUGH SOCIAL MEDIA DISCOVERY

The advances in the IT sector and especially the Internet and Web 2.0 domain have led to a new digital era, era molded by the enormous volume of data generated (Big Data). Helped by the new and attractive ways of information creation and sharing, users are producing new content or constantly shaping and evaluating content created by other contributors. In this way many users surpass the margins of being just consumers and become authors, co-authors or promoters of the always-evolving social content. Led by the social media sites and tools and the new mobile ways of information exchange, the data generation and consumption is at its highest. Vast number of organizations, including educational institutions use the social media tools to their advantage and goal achievement. They even invent and introduce new ways of user collaboration and content consumption and sharing (such example are the Massive Open Online Courses - MOOCs). Therefore, the content shared at the social media sites should be scrutinized and evaluated for potential educational significance. By introducing a platform that will serve as integrator and as knowledge extractor and evaluator, this paper concludes and evaluates the findings of such implementation.
Furthermore this paper defines the access to the social media and the vast requirements of the different sources. It presents workarounds to bypass some of the limitations of accessing the social media and introduces particular algorithms in achieving these goals.
In it we introduce a new novelty metric: Learning Value Index (LVI) to rate the educational validity and relevance of the retrieved and suggested content. We also present algorithms for the LVI calculation in parallel with the different methods that affect the LVI final value as well as connectivity to classic Learning Management Systems (LMS) and the information exchange between these two systems. We analyze the notion of the “social fragments” and “sub social fragments” corresponding to class or book chapters and present a piloting reports taken from approximately 100 users and 2 experts.