The dissertation is devoted to solving the scientific and practical problem of increasing management decision-making efficiency by using models and methods of differentiation of consolidated information data.
The purpose of the dissertation determined the need to set and solve several specific research problems:
to analyze existing models and methods of consolidated information data differentiation on the example of the construction of the social portrait, to justify the choice of research objectives;
to develop fuzzy criteria to determine the significance of the social profile information components;
to develop information models of personal and group social portraits formation based on OSINT-technology of legal obtaining and use of information from open sources;
to improve the method of modeling the social profile using the optimization of the transformation of Big Data structures;
to investigate the developed models and methods of forming social portraits and develop practical recommendations for their application.
The object of the paper is the processes of consolidated information differentiation.
The subject of the paper is models and methods of consolidated information data differentiation.
Currently, decision support systems are widely used in various spheres of human activity. They need to process large amounts of consolidated information to function. Such processing often has to be done under time constraints and using insufficient computing resources of decision-making information systems. Decision-making systems mainly operate in real-time, which imposes stringent requirements on these systems' timeliness.
The management decision-making speed increase may be obtained by improving the methods of data processing management. Based on the consolidated information characteristics, data differentiation methods, which allow the processing of large amounts of data, are used to reduce data processing time. However, the currently existing models and methods of data differentiation do not allow to fully obtain the proper values of the operability indicators. Therefore, there is a contradiction between the increasing volume of processed information and requirements for speed of decision-making and existing models and methods of processing consolidated information, which determines the relevance of the purpose of this dissertation research
The dissertation research was conducted using the mathematical apparatus of graph theory, database theory, and the concept of non-relational data warehouses, Big Data technology, text analytics technology, parallel data processing methods, methods of research and construction of neural networks, multimedia data analysis methods.
To achieve the aim of the dissertation, the following research results were obtained: an information model of personal and group social portraits was developed, based on OSINT-technology, which allowed reducing the risks of confidentiality violations while maintaining the completeness and quality of the input data; the method of management of social processes has been improved, which differs from the known by the synthesis of the general system model of management into a single triadic hierarchical system, which allowed reducing the uncertainty in solving the problems of selecting different options for management decisions; the model of the process of determining the significance of the social profile parameters has been further developed, which differs from the known ones by using an improved coefficient of significance, which reduces the uncertainty of the initial data and increases the speed of decision-making; the methods and models developed in the work are a scientific and practical basis for the differentiation of consolidated information, and the algorithms and programs developed on their basis allowed reducing the time spent on the analysis of large amounts of data while maintaining the level of objectivity of the evaluation compared to the involvement of experts.
The practical significance of the results is that the methods and models developed in the work are a scientific and practical basis for the differentiation of consolidated information. The algorithms and programs obtained on their basis made it possible to reduce the time loss for the analysis of large amounts of data while maintaining the level of objectivity of the evaluation by 7% compared to inviting experts.
The developed Big Data processing models allowed carrying out the transformation of structures of large data sets by 5% faster compared to classical models.
The model developed on the basis of OSINT-technology has reduced the time to find the necessary information by 8%.
The dissertation research results were used in the work of Horizont LLC Kharkiv, Radiant LLC Kharkiv, Teploenergosistemy LLC Kharkiv and in the educational process of the EC department of Kharkiv National University of Radio Electronics.
