The aim of this study is to prepare a module and to test the effectiveness of the module which categorizes the questions with respect to the cognitive levels, determines the difficulty levels by calculating the related data of item difficulty and item discrimination, automatically tunes the difficulty levels of the questions with reference to the answers given in the process of the exam, and provides person-specific exams both in the classroom and digital environment. The Web-Based Expert Examination System developed within the scope of this study makes it possible to prepare questions compatible with the assessment and evaluation criteria in education enabling the measurement of the level of actual learning of the students. By utilizing the prepared questions, it would be possible to generate exam forms. This system records the data belonging to the students and the data as a result of the student-system interaction, and can determine the operation order of the system and the questions to be used to evaluate the knowledge level of the student in the light of the information obtained from the records. In addition, this system can improve itself according to the changing needs and student characteristics in due course and adapt itself to the current situation. A questionnaire form was developed by taking the measurement and evaluation criteria in education and qualities which testing tool was required to have into consideration. Quantitative analysis of the research was completed by utilizing the packet program “SPSS 21 for MacOS X” and items were examined via dependent sample variant analysis. The results of the study indicate that there was a significant difference between the referee evaluation scores in favor of the exam form, which was prepared by using the Intelligent Exam System in which Expert Agent and individuals interact, comparing the other exam forms according to the qualities that a testing tool is required to include.