Analisis dan Deteksi Kemiripan Teks Berbasis Python dengan Algoritma Levenshtein Distance

Improvements in information technology have complicated the issue of plagiarism in academia, particularly in higher education. This project intends to create a plagiarism detection tool that examines the similarity of PDF files to established references utilizing the Levenshtein Distance method. The suggested system can effectively and precisely identify plagiarism by a series of procedures, such as text extraction, linguistic Preprocessing (tokenisation and stopword removal), and calculating the degree of similarity using the Levenshtein Distance method. Testing was carried out on various scenarios, including variations in document size and plagiarism levels. The experimental results show that the higher the level of similarity between the document and the reference, the longer the computing time required. However, this system can detect plagiarism with a fairly good success rate, even in documents with a low level of similarity. Black box testing confirms that this application can work according to the expected specifications, namely inputting PDF documents, detecting plagiarism, and providing accurate similarity percentage results. This research contributes to providing a plagiarism detection tool that can help maintain academic integrity, with the possibility of further development through integration with machine learning and user interface improvements.