This paper studies the use of a multi-objective clustering approach for movies in the long tail. The long tail of movie datasets consists of niche films that have been overshadowed by mainstream movies and blockbusters. Recommendation systems have a golden opportunity to enrich the user experience by focusing on the long tail and serving customers with more personalized recommendations. Therefore, in this work, we use the multi-objective clustering approach to the movies in the long tail to take into consideration multiple objectives that are often conflicting in order to generate clusters that enhance content discovery in the long tail. We then predict the ratings based on the newly generated clusters by applying different clustering algorithms to generate the initial population and applying various objective functions in the clustering process.