There are an increasing number of research papers getting published day by day. It becomes difficult for a researcher to closely examine all the research papers in their research field and find out the papers that are related to their research work for guidance. Recommender system helps the researcher by recommending papers based on the ratings given by other researchers in that field. Collaborative filtering is one of the most successful technologies for building recommender systems and is extensively used in many commercial recommender systems. Unfortunately the computational complexity of these methods grows linearly with the number of users and number of items, which in typical research paper domain can be several millions. To address these scalability issues, we present an effective recommender system based on the subspace clustering, which analyzes the researcher-paper matrix to discover the relation between different researchers and uses these relations to compute the list of research papers to recommend.