AUTOMATIC METRICS AND EXPERT EVALUATION OF TRANSLATION IN A LANGUAGE PAIR WITH LIMITED RESOURCESMaqsudaHaitovaauthorThe article examines the reliability of machine translation quality evaluation metrics for the Uzbek–Russian language pair. A test set consisting of 600 segments representing three functional styles was compiled as the research material. The results of seven automatic metrics were compared with expert evaluations based on MQM. The analysis resulted in a reproducible evaluation protocol for determining the degree of agreement between automatic metrics and expert assessments.2026-01-01Article