试卷识别是智能组卷系统中数据库建立的基础,如何正确有效地识别出试卷中的题干部分并将之入库是一个尤为关键的问题.提出了基于SVM(support vector machine,支持向量机)的大规模试卷识别方法,该方法有效解决了进行题干识别之前大量人工预处理工作的问题.根据SVM原理,建立一个超平面,将试题文本分为两类——题序号和题干,并以题序号作为各题的起止点,进行试题识别.大量的实验显示,该方法具有可行性、有效性和适应性.经测试,识别率达100%. Paper identification is the foundation of the database in the intelligent test paper system, and how to correctly and effectively identify and store the test paper in the test paper is a key issue.This paper proposes a support vector machine (SVM) Based on the principle of SVM, a hyperplane is established, and the test questions are divided into two categories - title number and title and The title number is taken as the starting and ending point of each question, and the test questions are identified.Numerous experiments show that the method is feasible, effective and adaptable.It has been tested that the recognition rate is 100%.