Thursday, April 23, 2009

Attributor test delivers so-so results

Attributor pretty much represents the state of the art these days in tracking down copyright violators. To see how well it works, I took a test drive of the free version it offers to bloggers. The results are fair to middling.

Attributor works by gathering all the content on a publisher's website and then crawling the web to see if it finds sites that have picked up some or all of an article. When it finds a match, Attributor tells you what percentage of the article was used and links back to the site that used the article. In some cases, the quote is highlighted in yellow on the site that used it.

In a test of the free version offered to bloggers, which presumably is less robust than the version sold to major publishers, I discovered many new places where Newsosaur had been quoted that I had not known about before.

Most of the web publishers using Newsosaur content used limited snips from this blog and properly linked back to me. Many of the sites carried no advertising and none of them appeared to be large enough to generate sufficient advertising revenue to make me want to get on the phone to demand my fair share.

As useful as Attributor was, it also missed a number of places where my articles had not only been cited but largely reprinted intact. I knew about the prior instances in two ways: By tracking back inbound links to my blog and through Google agents that alert me when “Alan Mutter” or “Newsosaur” is discovered by Google as it trolls the web.

Attributor also reported a few false positives, including stating that some of my blog had been reprinted – which it had not – on the site of the consulting and investing firm where I am a partner.

Verdict: This is a workable but not foolproof solution. As Attributor grows and indexes more of the web, its hit rate presumably will get better. The product today will require a good deal of human intervention to be useful to any publisher.


Blogger Mike O'Donnell said...

Alan, I invite you to take iCopyright Discovery for a test drive. Rather than having our own crawler and indexing the web like Attributor, we use Google and Yahoo to find sites that are using the content, but only those that meet pre-defined criteria. I'm not saying Discovery is "better" because that is relative and depends upon the objectives of the content owner. It goes about the problem is a different way.

11:31 AM  

Post a Comment

<< Home