Machine Learning-Based Topic Classification of Economic and Financial Texts in Uzbek

Abstract

This article examines the problem of automatic classification of economic and financial texts in the Uzbek language. A specialized dataset was created for the study, comprising 100,000 sentences distributed across eight thematic categories: banking and lending, taxes and public finance, macroeconomics, financial markets and investments, currency and foreign trade, business and economic sectors, consumer prices and household economics, and other texts. Classic machine learning methods combined with the TF-IDF text representation were used as the primary approach. Four models were tested: Naive Bayes, Logistic Regression, Linear SVM, and CatBoost. Experimental results showed that the TF-IDF + Linear SVM model achieved the best quality, achieving an Accuracy of 0.914 and an F1-score of 0.914 on the test set. Additionally, the model's robustness was tested on texts covering other topics. The obtained results show that the proposed approach can be effectively applied for thematic classification of economic and financial texts and automatic filtering of irrelevant materials.

More by Nilufar Abdurakhmonova

• A model of a universal linguistic knowledge graph for Turkic languagesArctic XXI century · 2026
• Building a Phraseological Ontology of Uzbek Idioms and Evaluating Similarity Measures for Their Detection in TextInternational Conference on Computer Science and Engineering · 2026
• Developing Morphological and Syntactic Analysis Models for Karakalpak as a Low-Resource Turkic LanguageInternational Conference on Computer Science and Engineering · 2026
• WordNet as a Model for Developing a Bilingual Thesaurus of English and Uzbek Qualitative AdjectivesInternational Conference on Computer Science and Engineering · 2026