Model of an Intellectual Information System for Recognizing Users of a Social Network Using Bioinspired Methods

摘要

The analysis of social networks is one of the actively developing areas of research for today. In addition to methods of modeling social relations, the interest for researchers is also the methods of obtaining initial data. The analysis carried out by the authors showed that, as initial data, analysts use profile information, various types of activities. Consideration of photo albums and video files as a source of data on users of social networks was not considered. At the same time, by identifying the user's face and further searching it in photographs and video files in other profiles, an expansion of the list of basic data available to the analyst is achieved. The problems associated with the recognition and identification of the person in the photo and video are well known and not fully resolved. The greatest difficulty here is the problem of scene variation, since the angle and location of the faces in the photo and video cannot be known in advance. The article presents the results of solving the problem of scene variation, presented in the form of a model of an information system using bionspirited methods. This system and the method incorporated in it are intended for preliminary image processing in order to broaden the range of permissible scene variations.