Elective Learning Objects Group Recommendation Using Non-Cooperative Game Theory

Recommendation systems are applied to various domains to suggest appropriate recommendations to users in their respective domains. Group recommendation systems are used to offer items to a user group. Universities offer learning objects in each semester where student's background, expertise and interest are not fully considered. In this paper, we focus on designing a recommender system which will recommend a set of learning objects to multiple students. We are using group recommendation system to deal with the issue of recommending learning objects to a learner group. This study focuses on student's preferences to recommend a set of learning objects offered by the university rather than to some specific learning object. The system process each learner's preferences to set of learning objects from two perspectives, learner attraction to the learning objects and considering multiple factors such as student background, expertise and interest to find similarity among learners. This paper will enhance student learning objects selection process by considering different factors (interest, skills and learning object) which are essentials for selecting and offering learning objects in university. For finding student similarity we analyzed multiple factors of student and learning objects profile such as student interest (rating), skill assessment and course evaluation. Multiple factors with continuous values are used in cosine similarity technique to find similarity among the profiles of students. Finally, to deal with the issue of multi-decision group recommendation, we model the recommendation process as a non-cooperative game to achieve Nash equilibrium and demonstrate the effectiveness of our proposed model with a case study experiment.