TURIZM PLATFORMALARIDAGI O‘ZBEKCHA SHARHLARDA SEMANTIK MA‘NO VA SENTIMENTNI ANIQLASHNING SUN‘IY INTELLEKTGA ASOSLANGAN KOMPYUTER-LINGVISTIK MODELI, AN ARTIFICIAL INTELLIGENCE-BASED COMPUTATIONAL LINGUISTIC MODEL FOR IDENTIFYING SEMANTIC MEANING AND SENTIMENT IN UZBEK-LANGUAGE REVIEWS ON TOURISM PLATFORMS, КОМПЬЮТЕРНО-ЛИНГВИСТИЧЕСКАЯ МОДЕЛЬ НА ОСНОВЕ ИСКУССТВЕННОГО ИНТЕЛЛЕКТА ДЛЯ ОПРЕДЕЛЕНИЯ СЕМАНТИЧЕСКОГО ЗНАЧЕНИЯ И ТОНАЛЬНОСТИ В УЗБЕКОЯЗЫЧНЫХ ОТЗЫВАХ НА ТУРИСТИЧЕСКИХ ПЛАТФОРМАХ

<jats:p>This article presents a methodology for developing an AI-based computational-linguistic model for detecting semantic meaning and sentiment in user reviews on tourism platforms across three languages — Uzbek, Russian, and English. The study integrates transformer-based multilingual language models (multilingual BERT, XLM-RoBERTa), ontology-driven semantic analysis, and multi-class sentiment classification. Experiments conducted on a monoliguial tourism review corpus demonstrated that the model achieves an overall accuracy of 91.4%. The research results are of significant scientific and practical importance for Uzbek NLP development.</jats:p>