Data mining and an application in educational sector

Data storage capacity has increased thanks to today's increasing technological capabilities and within increasing volumes of data, it becomes more difficult to access meaningful information. Data mining techniques makes predictions about future by extracting the available complicated dataset containing useful information and using it. Many field such as marketing, business, electronic commerce, medicine and education are utilized data mining techniques. This research consists of the operation and details of the data mining processes, detailed information about the models used in the data mining, information about the package programs used in the research and application sections. The data used in the application section is the term-end evaluation survey which a state university applies to its students. The survey was applied to 5820 university students. Using SPSS Clementine and WEKA programs, logistic regression and decision tree algorithms were constructed on the data and the obtained results were interpreted. According to the analysis results, it was seen that the determinant question for the students was question included "the exam, projects and quizzes assisted to learning". However, it was determined that the students tend to answer the questions according to the course repetition criterion, that is, they found useful it if the lesson is passed when student take it first time. The other determinant question is whether the expectations at the beginning of the period are met and the sources are up to date. Accordingly, it was seen that the success of the courses is associated with the enhancement of the professional development of the student and the development of the worldview. On the other hand, the information obtained from decision-making structures also indicates that students tend to give close answers to questions. This situation also limits the information that can be obtained from the courses and the teaching members. For this reason, it is necessary the paying attention to these negatives while conducted the questionnaire. In addition, to increase the success of students and faculty members, it was thought that survey application time should be changed. Accordingly, survey shuld be appiled to students before the final exam, which will provide more detailed informaton. Keywords: Data Mining, Educational Data Mining