Plagiarism Checking Information System for Final Assignments of Malang State Polytechnic Using the Latent Semantic Analysis (LSA) Method

This study aims to design and develop a plagiarism checking information system for students' final assignments at the State Polytechnic of Malang using the Latent Semantic Analysis (LSA) method. Plagiarism in scientific papers is a serious problem that can damage academic integrity. Therefore, an effective and accurate system is needed to detect plagiarism in students' final assignment documents. The LSA method was chosen because of its ability to analyze semantic relationships between words and documents, thus detecting plagiarism that cannot be detected directly by word-based methods. This system was built to check the similarity between uploaded documents and documents in the campus database. Based on the test results, this system demonstrated a plagiarism detection accuracy of up to 80%, as well as an efficient processing time of between 13 and 14 seconds per document. Thus, this system is expected to improve supervision of the authenticity of students' final assignments and support academic integrity at the State Polytechnic of Malang. Recommendations for the development of this system are to expand file format support and improve the ability to detect more hidden plagiarism, such as paraphrasing.