MorphUz: Morphological Analyzer for the Uzbek Language

Abstract

The Uzbek language is an agglutinative language in that words are derived from stems (root) by concatenating affixes. This property makes a large number of combinations of morphemes, and greatly increases the word-vocabulary size. Therefore, words are split into certain sub-word units and applied to text and speech processing applications. Proper sub-word units not only provide high coverage and smaller lexicon size, but also provide semantic and syntactic information that is necessary for downstream applications. This paper discusses a morphological analyzer tool for natural language processing and machine learning purpose. The tool named MorphUz, which can split a text of words into a sequence of morphemes. Morphological analyzer is one of the main part of the natural language processing. MorphUz analyzer is an open-source morphological analyzer for the Uzbek language. The MorphUz analyzer is available as a website for exploration. MorphUz analyzer implements the morphology of the Uzbek language following a two-level approach using stemming and suffix analyzer. The implementation of MorphUz analyzer done by using PHP and JavaScript scripts and MySQL database.