THE METHOD OF DETERMINING PARTIALLY MEANINGFUL SYNONYMS IN IMPROVING THE UZBEKI THESYNOMYSER SOFTWARE

Abstract

In world experience, experiments are being conducted with the help
of many models and methods such as PMI, Edmond, DVS models, n-gram,
skipgram, and anticollocation methods to determine near synonyms. It is known
that determining synonyms based on substantive classification is a more
complicated process than determining formal synonymy. This is especially
important when creating an Uzbek synonymizer program. This article focuses on
the problems encountered in identifying the phenomenon of synonymy based on
substantive classification. Also, in order to eliminate such problems, the solutions
that are currently widely tested in the world experience were looked at and a
summary method was proposed.