This article is devoted to the problem of classifying Uzbek-language texts based on intelligent text analysis technologies. The study analyzes probabilistic approaches to text classification — Bernoulli and Multinomial Naive Bayes models. Text documents obtained from the official information source of the National Information Agency of Uzbekistan were used in the research. For the experiments, 600 documents belonging to 6 different categories and containing a total of 169,205 words were selected, and a comparative analysis of the effectiveness of the models was conducted using these texts.
| Mualliflar | Дадаханов, Мусохон |
|---|---|
| Jurnal | Al-Farg'oniy avlodlari |
| Nashr sanasi | 2026-03-13 |
| Son | 1 |
| Betlar | 191-196 |
| Til | Rus |
text, classification, probabilistic model, Bayes, Bernoulli model, multinomial model, NLP., текст, классификация, вероятностная модель, Байес, модель Бернулли, мультиномиальная модель, NLP., matn, tasniflash, ehtimoliy model, Bayes, Bernulli modeli, multinominal model, NLP.
This article analyzes the efficiency of the MQTT protocol in transmitting sensor data within IoT systems operating under constrained energy and network resources. The conducted experiments and developed mathematical…
ABSTRACT The operational complexity of modern industrial processes demands control frameworks that are both mathematically rigorous and computationally agile. This paper provides a systematic review of the…
Today, mathematical modeling is becoming more important in studying digital signal processing and image reconstruction problems. In modern information technologies, spline functions often give better results than…
Abstract: This article presents a comprehensive analysis of the influence of various modifiers and alloying elements on the mechanical properties of manganese steel. Based on the obtained results, a multi-component…
This study analyzes risks associated with different instructional approaches in higher education and explores mitigation strategies. With the growth of online and hybrid delivery, it applies Fuzzy TOPSIS to evaluate and…
In this study, an algorithm for artificially expanding the training sample was developed for classifying cotton varieties recommended for planting in the Republic of Uzbekistan by region. The algorithm is based on…
This paper investigates the construction of a quadrature formula based on B-spline functions for the approximate evaluation of definite integrals. In the study, the integrand is approximated using B-spline basis…
This article presents the results of scientific research on the development of a web-based software platform designed for the early detection of diabetes mellitus. The system, called “Erta Diabet”, has been developed as…
Abstract. This article proposes a firewall model based on dynamic filtering rules. The model represents packet processing as a single-server queuing system with a finite buffer under a Poisson arrival process and…
This article elucidates the principles of synthesizing state observers used for controlled objects with limited direct state measurement capabilities. The observer calculates the internal states of the object based on…