A High Performance Recommendation Framework For Online Learning Groups
For an online learner, they require a learning process and methodology that is flexible. The online learning system should be able to recommend suitable materials to a learner in a live environment. Much of the prevailing systems for online learning resource recommendation are working based upon collaborative filtering techniques. Such techniques have limitations when it comes to adapting to real-time and need the previous history of learners.
Hence, this study proposes to present a live learning resource recommendation methodology, appropriate for flexible and complex setups. The presented system is based upon reinforced learning challenges. The existing system explores the present environment to obtain data and exploit the information to make a decision. The proposed system will depend on live and changing data. The system being proposed is an enhanced recommendation method, which presents learning resource recommendation by obtaining student’s behavioural information. Firstly, it will create user clusters based on their learning styles. Secondly, it enforces association rule mining to obtain all the choices and behavioural trends of each individual cluster. Finally, it will generate customised reference sets of differing sizes. The proposed system is a recommender methodology that references online materials for students towards their learning efforts depending upon the sameness of student’s learning history and data mining approaches to uncover similarities between courses
Keywords: smart learning; expectation maximisation algorithm; e-Learning; apriori algorithm; online learning methodology; adaptive recommendation