Evolutionary model of knowledge representation

Abstract

This article focuses on an evolutionary model of knowledge representation. This model is a self-reproducing knowledge system, capable of generating new knowledge based on existing models and user-defined data types through evolutionary computation operators, determining knowledge inference, and identifying the most useful information based on precedents and affinitive analysis rules. This model of knowledge representation was created to increase the degree of independence of intellectual systems by automating the process of knowledge generation about the domain of the problem being solved, rules of their application and reaction of environmental objects to actions made by the system, to increase the variability of solutions provided by the system, to increase the information content (by changing the classical tree-like structure into an evolutionary one) and to implement the possibility of creating new models on the basis of the parent models' main components. The concept of this model and its implementation may allow creating models of knowledge representation, which will not only store and describe knowledge and connections between them, generate new knowledge and perform independent optimization, but also solve the tasks of each of the model components separately or in aggregate, i.e. perform evolutionary calculations for optimization tasks, search and analyze similar and interdependent elements in data structures, and provide knowledge representation in computer systems.