Who would you listen to?: A Frequent Itemset Mining and Recommender System approach on the Million Song Dataset