TOPIC CLASSIFICATION OF UZBEK TEXTS

Topic classification of text is an important task in the field of natural language processing, where the main goal is to classify text data into predefined categories. In this study, we analyze the dataset creation and methods for evaluating multi-tag news series as part of topic-based topic classification. First, in the article, we present obtained corpus for the classification of texts in Uzbek language. This corpus was collected from 5 different news and press websites and contains 4 categories of news, press and legal texts. We also trained a rule-based machine learning model and a neural network model for subject classification in this newly created corpus experiment. Experiments show that models based on recurrent neural network and convolutional neural network outperform rule-based models.