Research And Application Of C Program Feedback Method Based On Spectrum Analysis

With the application of automatic code evaluation technology in the field of teaching more and more,many colleges and universities in China are working to develop a support system suitable for the teaching needs of the school.C language programming course as an introductory course for learning programming,most of the audience does not have the programming foundation and lack of the ability to debug,so the correction of code error is a common difficulty in students.However,the feedback on the black-box evaluation technology dedicated to the ACM system is limited.Users can only get evaluation results and grammatical errors,but can not effectively guide students to locate logic errors in the code.In the long run,the shortcomings of insufficient feedback will gradually ruin the students' interest in solving problems and weaken the benefits of test-driven learning.In order to make the automatic evaluation technology serving the competition better meet the teaching needs of the programming class and improve the current situation of insufficient feedback of the evaluation results,this paper conducts research work on the logical error feedback method of the C program.The main research and development work includes three aspect.Firstly,this paper studies the application of automatic error location technology in the direction of code evaluation,combines the spectrum-based lightweight positioning technology with the needs of C language teaching,then proposes a C-program feedback model based on spectrum analysis.By analyzing the positioning effect of the feedback model on the actual dataset of the Online Judge(OJ)system,it shows that the model can greatly reduce the scope of the error code review,and can provide effective guidance for students to find program logic errors.Secondly,the C-program feedback support model based on spectrum analysis is applied to design and implement a web-based C language experiment support system.This system can guide students to locate program errors faster and then correct the program.The online assisted teaching method based on automated evaluation and feedback provided by the system,on the one hand,realizes the automatic review of the experimental report,which enables the teacher to be freed from the heavy experimental report correction and greatly reduce the work pressure;on the other hand,diversified forms of feedback also greatly enhances the ease of use of the system and the student experienceFinally,this article addresses the security challenges that online evaluation brings to experimental systems.The paper combines the service environment and technical architecture in the system to improve the prevention of duplicate submissions at the source and application levels.We designed a security architecture based on static analysis,account control,operation debounce,token interception,and account privacy for security issues in the code evaluation environment.And we have strengthened the security flaws of the existing programming problem evaluation module.Our test verified that the security system effectively improved the security of online code evaluation.