A generalized concept of knowledge collection and processing technology for decision-making support under uncertainty

• Date: 2025
• Main Author: Каденко, С. В.
• Format: Article
• Language: Ukrainian
• Published: Інститут проблем реєстрації інформації НАН України 2025
• Subjects: decision-making support; goal-oriented hierarchic decomposition; rating; ranking; strategic planning; scenario analysis
• Online Access: https://drsp.ipri.kiev.ua/article/view/354611
• Journal Title: Data Recording, Storage & Processing

In this article the author suggests a generalized concept of information collection, transfer, and processing technology for decision-making support in uncertain environments. First, key features of uncertain environments and a generalized decision concept are outlined. Decision is defined as the choice of one or more of the existing alternative options, such as objects, projects, strategies, or scenarios. Second, existing approaches are analyzed in retrospect, including the analytic hierarchy and network processes, ELECTRE, TOPSIS, PROMETHEE\GAIA, and DEMATEL techniques. Special attention is given to technologies based on complex dynamic evaluation of alternatives. The key common features of the analyzed technologies are: 1) usage of expert data alongside data from other available sources and 2) construction of a knowledge base for a given uncertain environment in the form of a weighted hierarchy\network graph. Nodes of this graph represent criteria and factors, describing the chosen domain in the context of a specified main goal, set by the decision-maker, while edges represent relative impacts of these factors and their respective weights. Such a knowledge base allows decision-makers and analysts to solve problems of several types, such as: rating and prioritization of alternatives and projects, allocation of limited resources to projects within a specified period of time, strategic planning, and scenario analysis. Based on conducted analysis, the author suggests a general classification of outlined technologies according to information sources used, complexity of knowledge bases built, and types of problems solved. A separate section is dedicated to challenges and prospects of generative AI application for decision support in uncertain environments. The final outcome of the paper is the unified concept of information collection and processing technology for decision-making support in weakly formalized domains, characterized by high uncertainty levels, complete with references to respective mathematical and software tools that should be used at each stage of technology implementation. Tabl.: 1. Fig.: 1. Refs: 49 titles.