OPTIMIZATION OF TEST KEEPER SCHEDULING USING GENETIC ALGORITHM AT INFORMATICS DEPARTMENT PETRA CHRISTIAN UNIVERSITY

before mid or final exam, there will be a manual process to schedule the test keepers for every exam session. The test keepers are lecturer assistants (assistant is an appointed student to help lecturer in class). For an exam session, the keeper can be 1 up to 3 people, depending on the exams participant. These manual process is considering many factors, i.e. the assistants batch (year), the average of exams participant batch(year), gender combination of the keeper, evenness of the exam keeping of every assistant, the character of the assistant itself, and the exam schedule of the assistant. These factors are considered upon picking every exam sessions keeper, which is taking a lot of time and knowledge, and this process is done twice a semester by an exam coordinator (lecturer). In this paper, will be designed an application that is using genetic algorithm to automatically assign the test keepers for every exam. The result of the application is tested during the mid-exam and final-exam early semester of 2016, and the application is giving a good result, with the accuracy of 90.23%, in which the 9.77% is some minor changes that is required to make the test keepers more suitable.