A Corpus-based Study Of C-E Translation Of Abstracts In AI-Scientific Papers

Chinese-English(C-E)translation of abstracts of science and technology papers in the field of Artificial Intelligence(AI)is irreplaceable in the promotion of the academic communications of AI-research internationally.Nevertheless,there often exists translationese or problematic expressions in C-E translation of AI-abstracts by Chinese scholars,for instance,inappropriate expressions of technical terms and non-standard scientific writing.As a result,this thesis selects abstracts from international journals according to impact factor,builds a mini-Comparable Corpus of Abstracts of AI-scientific Papers(AIACC).Based on self-built comparable corpus,this thesis summarizes the normative English of abstracts of AI-scientific papers by English-speaking authors then takes it as an ideal English abstract norm.At the same time,this thesis selects Chinese and its C-E translation abstracts with the same theme by Chinese scholars,constructs a mini-Parallel Corpus of Abstracts of AI-scientific Papers(AIAPC),then analyzes the translationese or inappropriate expressions in C-E translation practices by Chinese scholars.Based on self-built comparable and parallel corpora,the linguistic features of normative English and C-E translation abstracts are both discussed and summarized from lexical,voice usage and readability levels.The findings show that there are translationese or Chinglish expressions in C-E translation of AI abstracts by Chinese authors,such as inappropriate application of pronouns and adverbs;excessive usage of passive voice and poor readability.This thesis analyzes the potential reasons behind these problems,then puts forward some suggestions to improve C-E translation of abstracts of AI-scientific papers by Chinese authors.To be more specific,learn from the use of pronouns and adverbs from normative English abstracts,appropriately apply "WE+V" structure and avoid the redundance of pre-modifiers and subordinate clauses.As a matter of fact,this thesis reveals the latest theoretical development of corpus-based translation studies.