G.M. Dovudov

Annotation:

This article describes a detailed description of an algorithm for automatic morphological analysis of Tajik words. The developed algorithm determines the part of speech of a word based on the part of speech of the root and an analysis of the prefixes and postfixes attached to it. A morpheme database consisting of 81 prefixes, 65,422 bases, and 2,847 postfixes was used in the study. The analysis process consists of four main stages and presented as an algorithmic flowchart. The developed algorithms were implemented in a software package for developing an automatic morphological analyzer of Tajik words. The logical structure of which is presented as a diagram. A large set of literary texts was processed using the software package. The effectiveness of the automatic morphological analysis and the accuracy of its results were found to be 92.02%. Morphological analysis and tagging of word forms are one of the key stages in the process of creating national text corpora of the language.
The obtained results are widely used in the development of national text corpora, as well as automatic spell-checking systems and other programs for processing texts in the Tajik language.
Keywords: Tajik language, morphological analysis, prefix, stem, postfix, flowchart, expert morphological analysis.