Nowadays, information systems protection against social engineering attacks is one of the main issues in the field of informational security. Intruders often refer to publicly available sources where users place personal information of various types. This information makes it possible for an intruder to determine the most effective way to attack user. In this paper we discuss one of the approaches to search and collect user's public information in an automated fashion on the example of social network VK.com. Besides, prototyping of the software for text information analysis for evaluation degrees of vulnerabilities included into the user's vulnerabilities profile was presented.