Description: The article is devoted to studying of social networks user’s profiles. An overview of the current problems in collecting and processing data from social networks is completed. The main problems in the automatic collection of data from profiles of Internet users are identified. They are confidentiality of data, poor data structuring, access restriction and blocking, the dimension of data. Often access to user data is only allowed for registered and authorized network members, which requires support for user session emulation using special accounts. In many cases, the social interface APIs have a limited functionality that requires support for receiving static copies of HTML pages with the user interface. In order to prevent unauthorized automatic data gathering and limiting the load on the infrastructure of a social network service, service owners often introduce explicit or concealed restrictions on the permissible number of requests from one user account and/or IP address per unit time counting the number of queries referenced. The dimensionality of the data necessitates a parallel method of data collection, as well as methods for obtaining a representative sample of users of the social network (sampling). An overview of the methods of data clustering obtained from social networks has been reviewed. The possibilities of using social network monitoring data for decision making have been analyzed. An analysis of existing software solutions was made and it was found that most software products only allow constructing of a graph of data from a social network. Software that allows analyzing of user profiles of social networks is not available on the market. The article deals with the design of such a software system and the requirements for it were developed.
Keywords: profile, social networks, data collection, clustering methods, software products