This paper describes the part of recommender system designed for movies' critics recognition. Such a system allows the automatic collection, evaluation and rating of critics and opinions of the movies. First the system searches and retrieves texts supposed to be movies' reviews from the Internet. Subsequently the system carries out an evaluation and rating of movies' critics. Finally the system automatically associates a numerical mark to each critic. The goal of system is to give the score of critics associated to the users' who wrote them. All of this data are the input to the cognitive engine. Data from our base allow making correspondences which are required for cognitive algorithms to improve advanced recommending functionalities for e-business and e-purchases websites. Our sesystem uses three different methods for classifying opinions from reviews critics. In this paper we describe the part of system which is based on automatically identifying opinions using natural language processing knowledge.