Digital transformation and graphic interpretation of social interests of university students

Makhnev Sergei Alexandrovich, Dekanova Nina Petrovna

The proposed work is the initial stage of a study devoted to the analysis of textual data obtained from social networks in order to identify relationships between graduates of various specialties of the university and incoming applicants. The semantic-graphical method of digital transformation and graphical interpretation of unstructured text data is considered. The basis of the semantic method is a combination of empirical and theoretical approaches to determining the amount of information K Shannon and Poisson's uniform distribution. The resulting characteristics make it possible to single out the most valuable social interests of the study participants. The use of trigonometric and exponential functions as guides in the graphical interpretation of weight coefficients is proposed. The totality of the obtained matrices of weight coefficients and raster images is the input data for the subsequent formation of a trained neural network designed to recognize the proximity of the applicant's social interests to the collection of interests of graduates of certain specialties.

Keywords:text element, semantic analysis, amount of information, weight factor, term set, set collection, graphical representation, bitmap.