A Report On Bilingual Terminology Extraction From Xi Jinping:The Governance Of China

Computer-aided translation tools are widely used in translation practice.They can not only improve the translator's work efficiency but also can ensure the high quality of the translation.Terminology management is an important part of the computer-aided translation and plays an increasingly important role in translation practice.Terminology extraction is the first step in terminology management.The efficiency of terminology extraction is related to the efficiency of the entire terminology management,which in turn determines the quality of the translation.Which bilingual terminology extractor can be used to extract terminology more efficiently relates to the quality of terminology database and the quality of the translation,and also relates to the improvement of translation efficiency of computer-aided translation tools.In response to this problem,this experiment uses the Chinese and English versions of Xi Jinping: The Governance of China as the basic corpus.Bilingual terms of the three important chapters(the third,fifth and seventeenth chapter)of this book are manually labelled as the �golden� data set for this experiment,and then the three bilingual terminology extractors,Multiterm Extract,Sketch Engine,and LingoSail TermBox are used to extract bilingual terms automatically.Then the result of extraction is compared with the result of manual annotation.By using the two parameters of accuracy rate and recall rate as benchmarks,the automatic extraction efficiency of the three bilingual terminology extractors is compared.At the same time,it analyses and compares the differences between the extraction efficiencies of these three terminology extraction tools for word-level and phrase-level terms,and analyses the features of the missing terms of the respective tools.The results show that the comprehensive extraction efficiency of LingoSail TermBox is the highest,which is higher than the other two software in both accuracy and recall rate.It shows that the extractor LingoSail TermBox can extract more bilingual terms,and the terms are relatively accurate.It will provide some reference for the translator's translation practice,and the experimental results will also provide some guidance for and contributions to the improvement of computer-aided translation software and work efficiency.