Implementasi Metode Perceptual Hash Untuk Deteksi Plagiarisme Tugas Mata Kuliah Software Modeling

Abstract

Plagiarism is a crime that is no stranger to the world of education in Indonesia. Plagiarism can be done in the form of text or images. One example of plagiarism in the form of images is when submitting assignments for the Software Modeling course at the STIKOM PGRI Banyuwangi Campus. One example of plagiarism that was carried out at the STIKOM PGRI Banyuwangi Campus was when a student copied a drawing assignment from another student. However, only changed in a few elements.Therefore, we need a mechanism that can detect plagiarism in the form of images.One method that can be used to detect the similarity of image sources is to use the perceptual hash algorithm. The algorithm is applied to compare one image with another image that is suspected of plagiarism. In applying this perceptual hash method, it requires a program using the PHP programming language.The results of this study are to detect plagiarism with an 8x8 image resize which has the fastest computing time of 0.437 seconds. When tested, the perceptual hash method was able to detect plagiarism with a success rate of 70.7% with a threshold value of 75% for the similarity level. However, it is not capable of images that change position. Thus, the perceptual hash method is capable of detecting plagiarism of sequence diagram images without changing position or normally.