Litcius/Paper detail

Artificial Intelligence Translator DeepL Translation Quality Control

Linlin Li

2024Procedia Computer Science19 citationsDOIOpen Access PDF

Abstract

With the advancement of language technology, artificial intelligence translators are also constantly developing. As an intelligent translator, the emergence of DeepL provides new opportunities for the development of human translation industry. Based on the theory of machine translation quality control, this article analyzed the translation quality control strategy of DeepL. By analyzing the performance of the designed source and target languages in the text input stage, the impact of part of speech tagging and syntactic analysis on translation quality in the text processing stage was evaluated. The machine translation quality evaluation method based on semantic similarity calculation was adopted to evaluate the translation quality of DeepL, and DeepL was compared and analyzed with other translators. This research found that DeepL translation performed well in terms of translation accuracy, fluency, and naturalness. The overall score of DeepL was 100, reaching 94.13.

Topics & Concepts

Computer scienceTranslation (biology)Artificial intelligenceQuality (philosophy)Control (management)Natural language processingChemistryBiochemistryMessenger RNAGeneEpistemologyPhilosophyNatural Language Processing TechniquesTopic ModelingSemantic Web and Ontologies