A NOVEL RESEARCH PAPER RECOMMENDATION SYSTEM

Research often spend a considerable amount of time searching for published papers and articles relevant to their interest, dissertation and research work. A recommender engine is a tool, a means to answer the question. “What are the best recommendations for a user?” Using trust in social networks provides a promising approach to make recommendations to other user based on trust propagation in finding research papers or research papers of a friend/research with similar interests. However, current recommendation algorithms are based on user-item rating. A collaborative filtering based research paper recommender system is proposed here with User and Item Based collaborative filtering approach to implement a recommender system for Research Paper. Users will enter details when they create a profile with the system. Based on the detailed profile and current contents of the recommended item in the database, the system recommends different research paper to user. User Based Recommender uses preferences of their other similar users for recommendation of research paper. Item Based Recommender uses User’s Item visit history for Recommendation. Four algorithms are implemented in each category by using Standard Dataset. Based on result generated, all methods are compared based on recommender accuracy in terms of Precision and Recall.