计算机编程语言阅卷流程工作任务繁重,评分效率不高,而且又过分依赖于阅卷老师的主观情感。如何能够科学地提高阅卷效率,对计算机编程类应试给予客观公正的评判,是目前计算机基础实验教学所关注的研究热点难点。通过静态语义理解思想的运用,结合程序依赖关系匹配度模型方法,将考生的作答结果标准化为程序系统依赖图集,并根据语义分析和匹配度度量方法予以判分。通过大量的有效实验数据分析论证,本文提出了一种基于程序依赖关系匹配度的C语言程序设计题评判策略,改进了常规阅卷智能化系统评分的不足。
The marking process using computer programming language is challenged with high workloads and less effective scoring.It also relies on teacher's subjective feelings.How to scientifically improve the efficiency of reading,give an objective and fair evaluation on computer programming class exam is a hot research topic.At the same time,it is currently the concerned difficulties.In this paper,we combined the static semantic understanding of the use of ideas with the program dependence relation matching degree model method.The examinees' answer results were standardized as procedures for system on atlas,and scores were given according to semantic analysis and matching.Through a large number of experimental data analysis,in this paper we presented a design evaluation strategy of C language program dependence based on matching degree,which improved the deficiencies of conventional scoring intelligent system score.