ПРОБЛЕМА РЕПРЕЗЕНТАТИВНОСТИ ТЕКСТОВ ДЛЯ ЭЛЕКТРОННОГО КОРПУСА УЗБЕКСКОГО ЯЗЫКА

Abstract

This article is devoted to the issue of text representation for the Uzbek language electronic corpus (http://uzbekcorpus.uz). Factors such as the statistical method, size, quantity of text selection for the corpus are one of the important criteria of the corpus design. According to this, the representativeness of the text is different for oral and written texts. Graphics of written texts, standardization requirements, adaptation of text elements for the accuracy and convenient of use of computer analysis, as well as transliteration of audio texts and assignment of segment units to the corpus manager in the appropriate order are the basis of text representativeness.