DETEKSI KEMIRIPAN SOURCE CODE PADA BAHASA PEMROGRAMAN JAVA MENGGUNAKAN METODE ANALISIS LEKSIKAL

Abstract

Programming Laboratory of Informatics Engineering Department of Engineering Faculty, Halu Oleo University is a learning place for college students majoring in informatics to follow programming class that become one of the standard to determine the student’s level of understanding in creating a program. Similarity detection method using Lexical analysis can be used to convert the programming code into an unique token. This token later used to identify the structure of a program that is being examined. After that, the token that already formed from two different source codes will be sorted alphabetically before entering the string matching process by using arrays. The string matching process will do the matching based on the position of each array. The more the similarities of the structure and the program codes, the more likely the code represented as plagiarism. The percentage value resulting from the token matching process later will be used as reference to determine the percentage of plagiarism in the program. If the similarities percentage greater than 70%, the examined source codes can be considered as plagiarism, but if it less than 70% the examined source codes will be not considered as plagiarism. Keywords— Source Code, Lexical Analysis, Java, Token, Plagiarism.