Our Anti-Plagiarism Module
In 2007 we have launched our own anti-plagiarism module - and it continues to work on our servers today. The module is separating the text into small blocks, called patterns. Then the system performs a web search by utilizing a so-called Rabin-Karp algorithm of string text search, thus giving the most precise results, (it even includes sources where the parts of the paper were copied from). This technology was invented back in 1987, and we found it to be an adequate solution for our clients' needs. The system is powered by such search engines as Yahoo!, Google, MSN.
However, the system has its disadvantages too: since this is machine-based recognition, it cannot recognize which part of the paper has been properly quoted and which has been plagiarized. All quotations, reference lists, and works cited pages will be displayed as plagiarized. The reason for this is simple – the plagiarism scanner only shows the match percentage, not the actual plagiarism score. Even properly quoted phrases will show a match on the report. For that reason, a score of under 20% usually raises no questions, leaving space to possible ‘search matches’. In case the papers scores higher than 20%, the paper is reviewed by our editorial team, and the paper is reviewed and analyzed for originality manually.