摘要
针对双语术语抽取系统在处理多种语言时大量耗费人力、财力、物力等的局限性问题,提出基于平行语料库的双语术语抽取方法,同时,分析此方法的不足,结合多种方法进行改善。在此基础上开发了一个双语术语抽取系统,并通过实验分析了相似度函数、语料规模以及改进后方法对术语抽取结果的影响,而且该系统已经作为商品化软件投入实际应用。
Since the bilingual terminology extraction system exists the limitation problems of consuming massive manpower,financial resources and material resources while treating with multilingual languages,the bilingual terminology extraction methodbased on parallel corpus is put forward. The shortcomings of this method are analyzed,and improved with a variety of methods.On the basis of the above,a bilingual terminology extraction system was developed. The similarity function,corpus scale,andthe effect of the improved method on terminology extraction results are analyzed with the experiment. The extraction system wasput into practical application as a commercial software.
作者
蒋俊梅
JIANG Junmei(School of International Studies,Henan Normal University,Xinxiang 453002,China)
出处
《现代电子技术》
北大核心
2016年第15期108-111,共4页
Modern Electronics Technique
基金
2014国家社科项目:基于语料库的学术英语元话语特征对比研究(14BYY150)
2013河南省哲社项目:学习者学术英语语篇衔接词语发展性特征研究(2013BYY024)
关键词
平行语料库
词对齐
共现矩阵
相似度函数
双语术语抽取
parallel corpus
word alignment
co.occurrence matrix
similarity function
bilingual terminology extraction
