Analysis of Approaches to Automation of Problem Structure Identification

The paper is devoted to the analysis of automatic and automated decomposition methods for their possible use in constructing methods for identifying the structure of dynamic weakly formalizable problems in artificial synergetic intelligence systems and, in particular, hybrid intelligent multi-agent systems. For this purpose, the paper presents the conceptual model of the problem, and analyzes five groups of automatic decomposition methods. A conclusion is made about the need to integrate existing decomposition methods and the prospects of the generative agent approach for developing methods for identifying the problem structure. The creation of such methods will significantly reduce labor costs for constructing synergetic intelligence systems for solving practical problems.