Developing a Decision Support System for Exam Scheduling Problem Using Genetic Algorithm

Abstract

Exam scheduling is a very complex process done every semester in every educational institution and is usually done manually. The limited resources in these institutions make the preparation of the exam program a demanding and inconvenient task. In addition, when exam scheduling is examined in detail, it is seen that it is a comprehensive task that requires fulfillment of many situations other than the appointment of the appropriate course for the appropriate time period. In this study, a solution approach that allows the assignment of courses to time periods is proposed for the educational institution whose data we use. Thus, it is aimed to obtain a program that can quickly solve the exam scheduling problem, which is a comprehensive task for the educational institution. A Genetic Algorithm, an artificial intelligence optimization algorithm, was used as a solution method, and the performance of the method on the problem was tested. In addition, taking into account students, lecturers, and administrative staff's performance within the study's scope and purpose, it aims to obtain exam programs that will satisfy everyone. However, most studies focused on the solution to the problem, and the idea of making this program suitable for the use of non-technical personnel was neglected. The program we developed in this study has been turned into a decision support system. Thus, the program has become a structure suitable for the use of non-technical personnel. As a result of the study, considering the institution's entire structure, a program that the institution can use in every exam period was obtained, and the exam program was automated, eliminating the time and effort spent by the institution staff.