In the 21st century, which is characterized as the Information Age, information access, knowledge quick learning is vital to the development of individuals and societies. With the use of technological innovations in the field of education in the information society, it will be possible to acquire a lasting place in the globalizing world. Distance Education refers to a model of the education system that students and teachers carry out through their learning and teaching activities by communicating via technologies and postal services. According to the year 2015 data, 68 out of 184 higher education institutions in Turkey offer an open and distance learning program. 47 of them are undergraduate, 17 are graduate, 11 are graduate completion and 56 are in master degree level. In total there are 505 different programs. The measurement and evaluation process of results of training is as important as developing content in distance education applications. When question types used in Distance Education Measurement and Assessment are examined, the use of Open-ended Questions is less than other methods. However, it is well-known fact that these type of questions are good predictors of the students' knowledge. The biggest problem that arises with the usage of open-ended questions is the evaluation part. The different answers that students will give to the questions, their personal narrative skills, or the interpretations they will answer in response to the questions make the evaluation process difficult. At this point, it would be better to interpret the answers recorded in the database with Text Mining methods and Natural Language Processing techniques. In this work, we implement an algorithm for evaluation of open-ended question. The experimental results showed that a correlation was found between 0,89 - 0,96 when evaluating open-ended questions of our system by teacher evaluation.