Plagiarism is hard to be avoid. Therfore, many system are invented to detect document plagiarism, like MOSS, TESSY, JPlag, CopyCatch, etc. The main idea to detect text plagiarism is by string matching. The algorithm used in this essay is Rabin-Karp algorithm. Rabin- Karp is superior in multiple pattern search. In this essay there will be some modification to improve Rabin-Karp performance. After some experiments were done, it turn out that modified Rabin-Karp has better results in running time than the general Rabin- Karp algorithm, while the generated similarity values results of the two algorithm are not far different.
Key words: text mining, string matching, Rabin-Karp, plagiarsm, stemming, hashing.