MODELLING NEURAL NETWORK SEGMENTATION OF THE MEDIA MARKET

Today, most Ukrainian media offices are not ready to face the realities of the market in the era of the modern information society and global digitalization: they don't know how to create high-quality content, how to promote it and where to find funding for its publication. The following organizational problems are relevant for Ukrainian media: a shortage of qualified specialists; combination of the post of chief editor, who is responsible for filling and subject materials with the position of CEO, whose tasks include the strategic development of the resource, the search for financing and other global tasks; lack of development strategy; KPI indicators are not used; the lack of a commercial department that would be engaged in the implementation of content. Modeling the regional television market based on intelligent data analysis methods is an urgent task. A large amount of data, their inaccuracy, incompleteness and uncertainty determine the use of ingenious methods, primarily neural networks and fuzzy logic. The article discusses the features of the development of regional media and television broadcasting, problems in the activities of local television channels. Also considered are the approaches associated with the development of the mathematical apparatus of economic research for modeling the regional television market, based on modern methods of data mining. The neural network information model of regional media market segmentation using Kohonen self-organizing maps allows revealing the hierarchy and complex structure of intra-system interactions of the participants of the territorial information space and substantiating the choice of the behavior strategy of its participants. The proposed model serves as the basis for a quantitative assessment of the structure of the regional media market and the characteristics of the formation of the target audience.