基于统计语言模型的英语易读性研究  

English Text Readability Based on Statistical Models

作　　者：邢富坤[1,2] 程东元[1] 

机构地区：[1]解放军外国语学院基础部,河南洛阳471003 [2]北京语言大学语言信息处理研究所,北京100083

出　　处：《解放军外国语学院学报》2010年第6期19-24,共6页Journal of PLA University of Foreign Languages

摘　　要：英语易读性研究有着悠久的历史,其研究成果在英语教学及研究中发挥着重要作用。近年来,随着统计自然语言处理技术的快速发展,英语易读性研究逐渐摆脱了过去单纯依靠测量词长、句长等因素的局限,开始从建立统计语言模型的角度进行探索,并取得了较好的应用效果。本文在回顾易读性研究历史的基础上,重点介绍当前基于统计语言模型的英语易读性研究的最新进展情况,并对这种研究方法进行评析,指出今后的努力方向,以期能够更深入地开展易读性研究,不断提高易读性研究水平。The study of English readability has witnessed a long history and played an important role in English teaching and research.In the past,the measurement of readability mostly depended on word length and sentence length,which fails to reflect the comprehensive nature of the understanding process.The measures are thus unreliable and the research leaves much to be desired.In the recent years,with the rapid development of statistical approach to natural language processing,text readability has been addressed from a new perspective and entered into a promising and fruitful stage.This paper reviews the new trends of readability study and also puts forward suggestions for future work.

关 键 词：统计语言模型 易读性 语料库 自然语言处理 

分 类 号：H319.3[语言文字—英语]