基于语料库翻译文体学的人工...能机译合同语言特征对比分析_.pdf

ABSTRACT

ContractsigningisacrucialaspectofChina’sforeigntrade,wherethequalityof

contracttranslationdirectlyaffectstheinterestsofbothpartiestoacontractand

influencesthesmoothprogressionofinternationaltrade.Withthecontinuous

developmentofartificialintelligencetechnology,manytranslatorshavestartedto

employmachinetranslationasanauxiliarytoolfordrafttranslationandpost-editingin

contracttranslationtoenhanceefficiency.However,fewstudieshaveanalyzedthe

stylisticperformanceofmachinetranslationenginesinthetranslationofcontracts.

Therefore,thisresearch,fromtheperspectiveofCorpus-basedTranslationStylistics,

comparescontractsgeneratedbyfourAI-basedmachinetranslationengines—BaiduAI,

ChatGPT3.5,TranSmart,andYoudaoAI—withthereferencetranslationcorpus

producedbyhumantranslators.Thecomparativeanalysisfocusesontexttypesand

specificlinguisticfeaturesthatreflectthestylisticcharacteristicsofcontracts,aiming

todeterminewhetherAI-basedmachinetranslationenginescanaccuratelyreproduce

thestyleofEnglishcontracts.

Tenbilingualbusinesscontractswereselectedasthedataforthisstudy,withtheir

human-translatedEnglishversionsextractedtobuildareferencetranslationcorpus.

ThentheChineseversionofeachofthetencontractswasfedrespectivelyintothe

aforementionedfourtranslationenginestoobtaintheircorrespondingEnglish

translations.Thesetenmachinetranslatedcontracts(Bystrova-McIntyre,2012;Sazzed

Jayarathna,2019)weresubsequentlycollectedasthebasisfortheestablishmentof

thefourparallelsub-corporaaccordingtothefourtranslationengines.Nelson’sBE06-

JservesasthereferencecontractEnglishcorpus,andBEC-Agreementsservesasthe

stylisticreferencecorp