Approach to assessing the level of intelligence of an information system

Abstract

Based on the analysis of the definitions of natural and artificial intelligence (AI), basic functions are proposed that reflect the cognitive activity of the brain. Their interpretation in a broad sense, using the rules of Russian grammar, made it possible to describe the entire range of concepts used in AI. The possibility of expressing basic functions in terms of simpler ones and combining them into more complex ones is shown. The completeness, consistency and non-redundancy of the basis of the cognitive functions of the brain are substantiated. The 14 basic functions are divided into 4 groups, reflecting different aspects of working with knowledge. This made it possible to consider the AI system as a specialized information processing system. For a sign that distinguishes the AI subsystem in the information system, a simulation approach to modeling the functions of the human brain is adopted. The level of intelligence of an information system is characterized by indicators of the share of hardware and software required for the implementation of AI models in the total amount of hardware and software of the information system. An example of assessing the level of intelligence of an information system of an unmanned aerial vehicle is given.