Abstract:

With the increasing frequency of scientific exchanges between China and Japan,the demand for patent documents and their translation has grown rapidly.In the context of the widespread application of machine translation,it is particularly important to explore the machine translation of patent texts.This study selected Google Translate,Wenxinyiyan (4.0) and ChatGPT (4.0) to focus on analyzing their translation of professional terms in the field of electronic communications.Combining the BLEU automatic evaluation index and multi-person manual.We compared the translation effects of the three machines,summarized the mistranslation features and proposed corresponding post-editing strategies.

Key words: patent text, Terminology Translation, machine translation quality assessment, post-editing strategy