Boris P. Korolkov

The article outlines the problem of constructing an artificial intelligence
system as a ranked knowledge processing system. Thermodynamic analysis of
nonequlibrium (with gradients) objects of different nature allows us to study their
evolution within the framework of self-organization. Their self-structuring that is
similar in many ways looks like a tree and it is described verbally (in terms of branch of
knowledge) forming its nomenclature. Introduction of numbers corresponding to the
numbers of ranks and the spots of specific names in the ranks leads to a digital model of
the evolving entity. Similarly it is possible to build up digital models for other branches.
At the initial stage of digitization of branches we can confine ourselves to threedimensional
matrixes of development consisting of cell-taxons. The principle of
uniform coding in multilevel rank structures of object evolution for standardization of
data base and knowledge base coordination within ontology approach has been
accepted. Thus a complicated diversified system reaches the AI level which technical
implementation assumes application of the system theory and system analysis
techniques.

Nonequilibrium thermodynamics, evolution tree, verbal and digital model, nomenclature, rank, matrix, knowledge base, ontology, artificial intelligence