Aplikasi Penilaian Jawaban Esai Otomatis Menggunakan Metode Synonym Recognition dan Cosine Similarity Berbasis Web

Examination is one of the evaluation tools in the teaching and learning process to determine the quality of student learning. But on the other hand, the exam is an additional activity that must be carried out by the teacher, especially in the correction process, because it takes a lot of time to match the results of the answers of different students, especially on questions with essay types. Therefore, a system is needed that is used to automatically correct student answers which is expected to help teachers to correct answers to essay questions quickly and easily. The method used in this research is a development method with a prototype model as a system development method and the Cosine Similarity and Synonym Recognition method. The test data used in this study were 116 student answers. The result of this research is a system that utilizes the Cosine Similarity and Synonym Recognition methods and has been tested using the RMSE accuracy test with several stages. The results of the RMSE test have a fairly high error value because there are student answers that are not similar to the answer key but the actual value gets a good score, there are students' answers that are typos, and there is a process that causes some terms or words to be missing, because the term it is included in the stopword list in the filtering process