Utilizing Lexicographic Resources for Sentiment Classification in Uzbek Language
Davlatyor Mengliev, Ergash Yu. Akhmedov, Vladimir Barakhnin, Zohid A. Hakimov, Oybek M. Alloyorov
Abstract
Nowadays it is seen the active development of text sentiment analysis tools, which help solve a huge range of problems, ranging from emotional tone or polarity to the nature of the analyzed text. Although significant advances have been made in sentiment analysis for languages with abundant computing resources, the problem still remains relevant for resource-poor languages, where Uzbek is one of them. The lack of complete linguistic databases and tools for this language represents a major obstacle for the scientific community.The present study aims to address this limitation by introducing a rule-based approach for sentiment analysis in the Uzbek language. In particular, the methodology uses three key vocabularies: the first includes a comprehensive list of more than 300 affixes used to generate word forms; the second focuses on cataloging exception words that defy standard morphological rules; and the third contains a list of word roots, some of the word roots are already marked with positive or negative sentiment polarity.When combined, morphological analysis with these lexical resources can effectively recognize tonal orientation in Uzbek sentences. This is especially important given the agglutinative nature of language and complex morphological structures that add layers of subtle meaning, influencing the interpretation of sentiments.