阅读理解系统是一种特殊的复杂问答系统,问题分析是重要的组成部分,对后续答题起着至关重要的作用。针对高考文学类阅读理解复杂选择题的选项分类进行研究,将选项看作短文本,对选项的类别标签进行了定义,引入依存语义和情感语义等多种特征对选项进行建模,使用多标签分类算法对选项进行分类,在语义层面上为阅读理解选择题的进一步解答奠定基础。
Reading comprehension is a complex question answering(QA)system,in which problem analysis is one of important components.Aiming at the college entrance examination,this paper studies classification of complex multiple-choice options in reading comprehension.We regard options as short texts and define their categories.Using the dependency semantic features and sentiment features,we build a multi-label classification model,which provides foundation for further question answer.