This is a long-overdue review of plagiarism detection systems (31 page PDF). The authors look at fifteen applications in multiple languages and evaluate them across a number of criteria including coverage and usability. They find numerous weaknesses and, with respect to range, find none of them suitable for academic use. They need to detect more types of plagiarism, indicate the source URL of the plagiarized item, and, they write, "Lose the single number that purports to identify the amount of similarity. It does not, and it is misused by institutions as a decision maker." And they emphasize, "Despite the systems being able to find a good bit of text overlap, they do not determine plagiarism." This is a meticulous report, well-referenced, and with full disclosure of methodologies and data.
Today: 5 Total: 85 [Direct link] [Share]
