AZIMUTHAL IDENTIFICATION SPHERE

Chochaev Alim Khuseevich, Tolgurov Takhir Zeytunovich

The article describes a special form of information representation within the boundaries of artificial intelligence. The authora reduce common forms of information organization to various variants of its matrix representation - from two-dimensional tables to multidimensional matrices - and note that in this form all of it is interpreted as a set of discrete unambiguous values - firstly. Secondly, the identification of any object in the environment of modern information representations complicates the typological recognition of the object by the machine, due to the need for a full cycle of attribution of such, even with an insignificant change in its parameters. The authors believe that the azimuthal identification sphere (AIS) is a universal architecture of information organization, which can be more functional than the existing forms of data "packaging". The article argues that AIS avoids the discreteness and unambiguity of information inherent in matrix methods of its representation and is effective not only in acts of object recognition, but is also effective in the processes of training AI systems and their classification activities in extensive thesauri.

Keywords:artificial intelligence, perception, attribution, information content, identification, cultural environment, azimuth, information module, archetype, matrix.