基于支持向量机的阅读理解试题难度预估研究  

Research On The Difficulty Estimation Of Reading Comprehension Tests Based On Support Vector Machines

作　　者：吴生蕾 任杰[2] Wu Shenglei;Ren Jie(Department of Humanities and Communication,Yangtze University College of Arts and Sciences,Jingzhou,Hubei,434020;Beijing Language and Culture University,Beijing,100083)

机构地区：[1]长江大学文理学院人文与传媒学院,湖北荆州434020 [2]北京语言大学语言测试和人才测评研究所,北京100083

出　　处：《考试研究》2022年第5期68-77,共10页Examinations Research

基　　金：北京语言大学院级科研项目(中央高校基本科研业务专项资金)资助,项目编号为21YJ140003。

摘　　要：试题难度反映试题质量,保证试题质量是保障考试信度和社会公平的关键。阅读理解试题是语言测试的考查重点,对阅读理解试题进行难度预估具有重要意义。支持向量机方法既可应用于线性可分数据,又可应用于非线性可分数据,本文采用支持向量机方法,以HSK(初、中等)阅读理解的第二部分试题为研究样本,对试题难度进行类别与数值的预估,分别以分类准确率、均方误差为评价指标。研究表明,支持向量机可用于阅读理解试题难度类别的预估。The difficulty of tests reflects the quality of tests,and ensuring the quality of tests is the key to guarantee test reliability and social fairness,while reading comprehension tests are the focus of language tests,so it is important to estimate the difficulty of reading comprehension tests.The support vector machines can be applied to both linearly and nonlinearly separable data.Therefore,this paper uses the support vector machines to estimate the category and value of the difficulty of the sample test questions by using the second part of the HSK(elementary and intermediate)reading comprehension tests as the study samples,and the classification accuracy and the mean square error as the evaluation indexes respectively.The study shows that support vector machines can be used for the estimation of difficulty categories of reading comprehension tests.

关 键 词：难度预估 支持向量机 阅读理解 

分 类 号：G424.74[文化科学—课程与教学论]