摘要
构建《数据结构》课程的领域本体和标准(学生)答案的句框架,在此基础上,形成一种新的主观题自动阅卷方法。该方法对标准(学生)答案进行分词、词性标注、句法分析、代词消解、提取句框架、计算词语相似度。与以往的主观题自动阅卷系统相比,该系统尽可能地让计算机以理解"语义"为核心,保持标准答案和学生答案的语义联系,评分准确率有所提高。
Constructs domain ontology for Data Structure course and standard （student） answer sentence framework, then proposes a new method of automatic marking Chinese subjective questions based on them. This method deals with the standard （student） answer in word segmentation, part-of-speech tagging, pronouns digestion, extracting framework, calculating word similarity. Compared with the traditional ones, this means allows the computer to understand the semantic information as much as possible, keeps the semantic relations between standard answer and the students＇, improves scoring accuracy.
基金
国家自然科学基金项目(No.61363036)
关键词
主观题
领域本体
句框架
Subjective Questions
Domain Ontology
Sentence Framework
