基于i+1方法的汉语文本难度测度  

Measurement of Chinese texts difficulty based on i+1 method

作　　者：周东杰 郑泽芝[1,3] ZHOU Dongjie;ZHENG Zezhi(Department of Chinese Language and Literature,Xiamen University,Xiamen 361005,China;School of Humanities,Fujian University of Technology,Fuzhou,350118,China;National Language Resources Monitoring and Research Educational Materials Center,Xiamen 361005,China)

机构地区：[1]厦门大学中文系,福建厦门361005 [2]福建理工大学人文学院,福建福州350118 [3]国家语言资源检测与研究教育教材中心,福建厦门361005

出　　处：《厦门大学学报（自然科学版）》2023年第4期670-678,共9页Journal of Xiamen University：Natural Science

基　　金：国社科重点项目“面向国际中文教育的常用词用法知识资源建设与智能信息检索平台研究”(22AYY014)。

摘　　要：针对当前文本难度度量通常将文本视为静态语言单位或抽象为语法结构,忽视文本语言单位组合难度度量的现象,提出一种基于i+1方法的汉语文本难度测度方法.该方法将N元字接续作为评测对象,通过利用不断动态“+1”的过程对文本难度进行判定评价,实现语言习得理论到文本难度评价上的迁移,从而对现有文本难度度量方法进行补充和改进.研究结果表明,该方法度量指标少、度量方法便捷,取得的文本难度度量效果相比于基于复杂特征的机器学习方法有明显提升.ed as grammatical structures,thus ignoring the measurement of combinational difficulty of language units in the text.To remedy this problem,we propose a difficulty measurement method of Chinese text based on i+1 method.By transferring language acquisition theory to text difficulty evaluation,taking N-character continuation as the evaluation object,and taking evaluation strategy of text difficulty determination as the dynamic process of"+1",we supplement and improve the current measurement research of text difficulty.In comparison with the machine-learning method based on complex features,our research result shows that the proposed i+1 measurement method of text difficulty with fewer indicators and convenient calculation secures better measurement performance of text difficulty.

关 键 词：i+1方法 N元语法 二元字接续 组合难度 文本难度测度 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]