Proposing Recommendation System Using Bag of Word and Multi-label Support Vector Machine Classification

Currently, recommendation systems (RS) have attracted a great attention from researchers both in academic and industry. They help extracting useful information quickly to provide right services to users in accordance to there basic requests which are commonly not too details. In the field of higher education and academic research, proposing project/thesis title that has already been developed to the right students/researcher for reference is a new and challenging issue. Therefore, we propose to build a system to suggest title of graduation project in the paper. In proposal system, we use natural language processing (NLP) that is applied artificial neural network (ANN) to solve feature extraction and training problems. The simulation results show that the proposal system achieves an accuracy of 82% with 12 s. The results also show that proposal system is suitable for applying in real environment.