Silalahi, Nardianas (2024) Sistem Pendeteksi Kemiripan Dokumen Text Menggunakan Metode Cosine Similarity dan Jaccard Similarity. Other thesis, Universitas Islam Riau.

Text 193510588.pdf - Submitted Version Download (2MB)

Abstract

The increased exchange of information through text documents in today's information age poses challenges in the management and analysis of large document similarities. Plagiarism, especially in an academic context, becomes a significant risk. This research focuses on analyzing the performance of Cosine similarity and Jaccard similarity algorithms in detecting the similarity of research proposal abstracts. While both have their merits, this research seeks a better understanding of when and how these methods should be used. The test results showed a relatively small difference in results between the two methods, with an average difference of about 3%. The data processing time also shows that the average difference between Jaccard similarity and Cosine similarity is not too far and tends to be almost the same with an average difference of 0.28 seconds with Cosine similarity being faster, the difference is small and acceptable. despite the difference in results, both are worth implementing in the Riau Islamic University Purse System to improve efficiency, accuracy, and integrity in the research process.

Actions (login required)

View Item