I'm guessing that the combination of "plagiarism finder" and "free" will light up the eyes of many who read this blog. I just learned of a new service called Plagium
. Using Yahoo search, it can take a string of text longer than will work in usual search engine front ends and look to see where on the web it might appear. I tested it with one paragraph from an article I had written for a magazine. Plagium turned up copies not only at the company's web site, but at Entrepreneur.com and Allbusiness.com. (Apparently the magazine sold the piece to Gale Group - yes, folks, the problem with magazines selling without permission to the database companies has not gone away, even if that long-standing class action effectively appears to have. Make sure you're registering your copyrights.)
I ran part of the paragraph through Google, which found only the Entrepreneur.com appearnance, and not the one at Allbusiness.com, so even in that one short experiement, Plagium appears to be the better choice.
At first I didn't enter the entire article, but just the one paragraph, thinking that looking for too great a section might identify potential copyright theft more easily than a long block of material, as someone might not have used the entire piece. But then, what if someone dropped the one graph I searched for? So I had Plagium search on all the text, which turned up the same instances.
So I tried something a bit trickier. When a piece I did on Wi-Fi hacking for the New York Times Magazine first came out, a good number of people posted it on various discussion sites, though I didn't know where it might appear any more. So I tried entering the submitted draft, and not the final copy-edited version. The only hits were - at the New York Times. So, Plagium will pick up examples that are close, but not exact.
Then I entered the exact text that the NYTM ran. Suddenly I had more hits, though the added ones were generally a few paragraphs with a link. If you prefer, you can choose to provide a URL for your article's location, rather than copying over the text. I think the latter makes most sense, because you reduce the number of false positives from other things that might be on the page.
Your search generates a graph in which potential infringements are bubbles on a timeline; the larger the bubble, the more likely that it is a copy. If you register with the site, you can have it look for new instances of the article over time, which suggests a smart set of steps:
- Write and submit the article.
- Finish edits.
- Register the copyright.
- Put the article into Plagium.
- Find the uses of the article as they happen.
Clearly this can't be your only tool, but it seems like a good one and, at least for now, is free. (Though there is a Donate button, which might be wise to actually use.)
If you're interested, here's a comparison
that the site PlagiarismToday (I can't get over that name) ran between Plagium, a paid infringement service called Copyscape, and Google itself. In this case, Google seemed to do far and away better, but the author says that there seemed to be a lot of duplication.
Labels: copyright, plagiarism