Core Insights - The article discusses the development of a new application by scientists at Surgut State University in Russia, aimed at analyzing machine translation errors to improve translation quality [1][2] - The application offers a more comprehensive analysis compared to standard methods, addressing the limitations of existing evaluation metrics [1] Group 1: Application Features - The new tool provides in-depth analysis of translation quality, focusing on vocabulary accuracy, semantic accuracy, and syntactic correctness [1][2] - It integrates multiple evaluation methods into a single automated tool, enhancing the efficiency of the analysis process [1] Group 2: Performance Analysis - The research team analyzed translations from mainstream online translation services and commercial neural networks, generating detailed reports for each translation [1] - Sentences with low scores in any evaluation metric are highlighted for further analysis, indicating areas for improvement [1][2] - While some translation tools performed well in vocabulary matching, all tested systems struggled with translating complex grammatical structures [1]