Abstract: Sociological surveys and tests exist for a long period of time and help to explore humanity to identify their interests and desires but we live in a world where almost everyone has their own profile in the social networks where they post pictures, write thoughts, notes and so on. Popular social networks in Russia can be represented as an online environment which is used to facilitate social interactions such as content sharing, points of view, experience and relevant information. In this study, we consider the method of support vectors for analysis of users’ profiles in social networks with the help of machine learning. We developed a system that could determine the selected network user opinion about the annexation of Crimea to Russia.