KOMPARASI ALGORITMA WINNOWING DAN ALGORITMA MANBER DALAM MENDETEKSI KEMIRIPAN TUGAS MAHASISWA

Komparasi Algoritma Text Mining

Abstract

Along with the rapid development of technology, the teaching and learning process, both carried out in schools and universities, has used information technology a lot to support the teaching and learning process. STIKOM Bali Institute of Technology and Business is one of the universities where most of the teaching and learning process uses information technology. In carrying out online lectures, several lecturers experienced problems, especially in correcting assignments collected by students. In some cases, student assignments are indicated to be almost exactly the same as other student assignments and some student assignments do not include the sources used. To overcome the problem of plagiarism, it is necessary to implement the matching technique of assignment documents in order to determine the level of similarity of student assignment documents. Many text mining algorithms can be used to perform string matching. In previous research, the author has compared the performance of the Rabin-Karp, Winnowing, Jaro Winkler algorithms with the results that the Winnowing algorithm has better performance between the Rabin-Karp and Jaro Winkler algorithms. In this study, the authors compare the Winnowing algorithm with the Manber algorithm in document matching. The result of this research is that the winnoing algorithm has better performance than the manber algorithm.