The dissertation is dedicated to addressing a current scientific and applied problem, which involves improving decision-making processes, developing scientific-analytical and methodological support, and creating components of information technologies for the rationalization of processes related to the provision and administration of social services to specific population groups.
The aim of the study is to enhance the efficiency of decision-making processes regarding the provision of social services to specific social groups in difficult life circumstances, under limited resources. Achieving this goal will ensure the efficiency, transparency, and adaptability of management systems for institutions providing social services, utilizing advanced information technologies.
To achieve this goal, models, methods, and tools have been developed to facilitate the selection of social service recipients, the distribution of services among providers, and the prediction of future needs.
For the first time, a fuzzy classification model and method have been developed, allowing for the determination of an individual's eligibility for social services based on minimal information.
An iterative method for distributing social services among providers has been proposed, which takes into account the priority of recipients and allows for the easy addition of new providers and recipients at any stage.
The method of decision reconciliation in multichannel decision support systems has been improved by developing new rules that account for the reliability of channels.
A decision support system architecture has been proposed that integrates various systems and data sources to optimize the processes of providing social services.
A method for designing semi-automated decision support systems has been developed, enabling effective management of multi-stage tasks in the social sector.
A hybrid forecasting method based on time series and expert assessments has been further developed, increasing the accuracy of social service needs predictions.
Methods for determining numerical assessments of objects have been improved through new rules for involving experts, enhancing accuracy in decision-making tasks.
The introduction to the dissertation justifies the relevance of the topic, defines the research objective, and outlines the main tasks to be solved to achieve this objective.
The first chapter of the dissertation systematizes aspects of the social service market functioning in Ukraine and analyzes data processing methods in the context of the study.
The second chapter focuses on modeling decision-making processes in the provision of social services under limited resources, as well as formalizing tasks at various stages of decision-making.
The third chapter of the dissertation presents the innovative content and targeted purpose of the main scientific-analytical results of the research:
A fuzzy classification method for selecting social service recipients. This method, in conditions of limited resources and a large number of potential recipients, allows for the effective and structured selection of those most in need and who meet the requirements set by service providers.
A step-by-step method for distributing social services among providers based on recipient eligibility indicators, ensuring that the distribution of tasks among providers meets the established criteria.
A hybrid method for forecasting social service needs, combining methods with expert conclusions. The method relies on a specialized "hybrid method block," which, using "expert competence levels (coefficients)," determines the resulting expert assessment of social service needs. By implementing two blocks (forecast and expert) of the hybrid forecasting method, two forecast values are obtained for calculating the final forecast value, proposed as a convex linear combination.
The fourth chapter is dedicated to designing components of decision support systems to support social service provision processes. The author justifies the principles of developing decision support systems for providing social services to the population. Rules for decision reconciliation in multichannel decision support systems with logical channels have been developed. A method and rules for decision-making in semi-automated decision support systems are proposed. This chapter also includes experimental verification of the obtained results.
The practical value of the work lies in contributing to the digitalization of social services, which enhances the efficiency of both state and private social programs.
The models and methods developed in the dissertation have already been implemented in several social institutions, improving decision-making processes and the distribution of social services in these organizations.
