Analysis and Evaluation of Terminology Translation Consistency in Scientific and Technical Literature

Abstract

In large-scale scientific and technical literature translation in which many people are involved, inconsistency in the translation of the same terminology is inevitable. Firstly, this paper carried out a comprehensive analysis to terminology translation inconsistency, finding that most are translations with same meaning but different indications, which influences the readability of the whole article. Then, we put forward a semantic similarity-based calculation method to identify this category of terminology inconsistency, selected translation with high frequency of network using Internet search engines, and carried out unionization. Finally, evaluate the literature improvement (post-editing) with two indexes of precision and consistency of terminology translation. In the experimental analysis, 100000-word patent document (English-Chinese human translation) is selected. The consistency index of original translation is 0.494, the consistency index of the processed translation is 0.763. The experimental result indicated that the method effectively improves the terminology translation consistency on the premise of correctly replacing terminology, thus significantly improves the readability of translation.