DETECTION OF PROFILE INJECTION ATTACKS IN RECOMMENDED SYSTEM USING MATRIX FACTORIZATIONA (DATAMINING) Item ratings play a crucial role in recommended systems to promote a particular product. Most of the customers know the quality of the product by reviewing its rating. If the product ratings are not genuine then the customer get the wrong details. This may decrease engrossment of the customer on a particular product. This is not feasible for a good recommended systems. In order to make the recommender system more robust, it is essential to identify profile injection attacks. Matrix Factorization is then employed to detect fake rating.