Experiment Report On The Evaluation Of English-Chinese Bilingual Terminology Extractors

Recent developments in computer-aided translation have helped enhance the efficiency of the translation process,of which terminology management plays a vital role.Establishing a termbase at the pre-translation stage helps translators maintain consistency in the project,ensure the accuracy of terminology translation and thus improves translation efficiency.Currently,multiple terminology extraction tools are available in the market,but choosing the suitable extractor has become a new problem for translators.In this experiment,a parallel corpus is built based on the original English version and the Chinese translation of The Troubled Empire: China in the Yuan and Ming Dynasties by the overseas sinologist Timothy Brook.The experiment uses three online English-Chinese bilingual terminology extractors currently available in the market,Tmxmall,Lingo Sail Term Box,and Sketch Engine One Click Terms,and two stand-alone terminology extractors,Synchro Term and SDL Multi Term Extract 2021,for term extraction.The purpose of this experiment is to evaluate the available extractors in terms of functional appropriateness,compatibility,performance efficiency,usability,and information security,as well as 11 subcharacteristics and 35 specific metrics based on the evaluation framework,in order to determine the best terminology extraction tool for textual materials about historical topics.By quantitatively measuring terminology extractors,the experiment aims to provide translators and translation teams with comprehensive information on selecting terminology extractors in a targeted manner.The experiment shows that among the five available bilingual terminology extraction tools in the market,the online tools perform better than the stand-alone tools in historical text.The stand-alone extractors are full-featured,but the terminology extraction needs to be improved so that they are moremore functionally correct,easier to start,and more beginner-friendly.Online tools perform better in Functional Correctness and Usability metrics.At the same time,the online extractors exhibit a discernible differentiation,which caters to both light users and core users with professional corpus application needs.In contrast,both stand-alone tools focus on professional users and overlook the light user market.Additionally,the experiment offers suggestions for further development of English-Chinese bilingual terminology extractors.