4th International Conference on Engineering and Natural Sciences (ICENS 2018), Kyyiv, Ukraine, 2 - 06 May 2018, pp.734-738
In this study, a genetic algorithm and a fuzzy logic based multiple choice test system is proposed. Learning is the process by which knowledge or skill creates long-term behavioral changes on individuals. Examinations are needed to measure the effectiveness of the learning process. These exams can be prepared by an instructor or automatically by means of a computer. Nowadays, with the widespread use of internet distance education has become popular, thus the necessity of creating an automatic exam through the internet has emerged. Although there are many methods for testing knowledge and skill acquired through learning, one of the most effective methods of measurement is multiple-choice test. The level of learning can be measured correctly by selecting the appropriate questions by optimizing the combination of the questions. Different attributes can be used in the selection of the questions included in the test to be created. Generally, the questions to be included in the multiple choice test are determined by the user determined difficulty, discrimination index and the frequency of the questionnaire. In the proposed system, the fitness function in the genetic algorithm consists of these attributes and is calculated using the fuzzy logic. It has been found that the system using the fuzzy logic is more successful than the system those not using the fuzzy logic when compared to each other.