EXAM SUBJECTS GENERATION FOR A COMPILER DESIGN DISCIPLINE DURING PANDEMICS
EXAM SUBJECTS GENERATION FOR A COMPILER DESIGN DISCIPLINE DURING PANDEMICS

Summary/Abstract: Education during pandemics was disrupted by the social distancing restrictions that were imposed by authorities. In this context, education was moved online together with all the knowledge assessment mechanisms like online examination. Oral online exams are good methods to examine students but it requires a lot of resources like time and tutors which usually universities cannot afford to allocate. Usually, exam results must be provided in a few days from the exam date and session periods are short. Written online examinations are prone to cheating due to the availability of multiple communication channels between students or students or third parties. Several cases were reported by tutors and they had to be discussed in the faculty’s management board. Expelling students from universities for cheating reasons may result in serious financial shortages. The only feasible way is to keep away students from cheating in online exams. One potential solution in this sense is to generate unique sets of subjects for each student and to allocate a limited amount of time for solving each subject. Thus, the solution of one student will not be reusable by others. On the other hand, the time limit will restrict the capacity of communication between students when supervised by video cameras. The video cameras must be at least two: one for identifying the student's face and the other focused on the written paper. To generate unique exam subjects, we propose an abstraction process to infer a template and a synthesis process where the template is instantiated with data computed from random numbers. The approach was tested on a group of 29 students and no cheating incidents were reported.

• Details
• Contents