Leveraging Natural Language Processing for Economic Insights: Sentiment and Key Indicator Extraction in Uzbek News Texts

Abstract

This article considers the problem of automated sentiment analysis and extraction of key economic indicators in texts written in the Uzbek language. The main goal of the study is to assess the emotional coloring of texts, as well as to extract quantitative indicators (for example, the exchange rate, budget parameters, refinancing rates), allowing for a more detailed analysis of the economic situation in the country. To achieve this goal, the authors developed a hybrid algorithm that includes a neural network module based on CNN + BiLSTM and a ruleoriented component for determining economic indicators. As part of the experiments, two datasets were collected reflecting both general economic texts and more narrowly specialized industry materials (agriculture, energy, banking). Testing results show that the accuracy (Precision) and recall (Recall) of the combined model reach $\mathbf{0 . 9 0}$, and the final F1-measure under optimal conditions reaches 0.93 . This approach allows not only to form a final tonality assessment, but also to provide analysts with a detailed report on the economic indicators mentioned in the text.

More by Nilufar Abdurakhmonova

• A model of a universal linguistic knowledge graph for Turkic languagesArctic XXI century · 2026
• Machine Learning-Based Topic Classification of Economic and Financial Texts in UzbekInternational Conference on Computer Science and Engineering · 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