An Approach to A University Recommendation by Multi-Criteria Collaborative Filtering and Dimensionality Reduction Techniques

Collaborative Filtering (CF) algorithms are most commonly used prediction technique in field of Recommender Systems (RS) for Information Filtering. It makes use of single criteria ratings that user have assigned to items which plays an important role in e-commerce to assist users in choosing items of their interest. For complex and massive dataset, Multi-Criteria Collaborative Filtering (MC-CF) frequently give better performance as well as accurate and high quality recommendations for users considering multiple aspects of items. CF algorithms need to be continuously updated because of a constant increase in the quantity of information, ways of access to that information, scalability and sparseness in rating matrix. Dimensionality Reduction techniques like: Matrix Factorization and Tensor Factorization techniques have proved to be a quite promising solution to the problem of designing efficient CF algorithm in the Big Data Era. This work aims at offering University Recommendation System, which combines MC-CF and Dimensionality Reduction techniques to provide high quality University/College recommendation to Students. The proposed solution not only reduces the computation cost but also increases the prediction accuracy and efficiency of the MC-CF algorithms implemented using the Apache Mahout framework.