Data Integrity

We've discussed metadata and data integrity issues before -- see especially this blog post for the drunks+lamppost analogy -- and have been advocating for better metadata standards generally.

Now Swivel launches a discussion of the authenticity of data that reminds us why this topic is so important. On their site, less than a third of the data is from an "official" source -- 69 percent is entered in by various site users. Hopefully, each user has entered in a data source along with the data, and I would hope no one would use Swivel as a primary data source anyway (any more than citing Wikipedia as a primary source.) Swivel is a great place to find interesting data and then chase it down to its source, but any site that aggregates other information shouldn't be your primary resource anyway.

The article on Swivel's blog points to Technorati Authority numbers as a way of looking at data reliability -- if lots of people cite it, it must be good, right? While I appreciate the attempt to point out an emerging rating standard on the internet (this blog's authority number as of today is 9, I was surprised to discover), the strength-of-numbers argument is remarkably not compelling.

So how do we establish a better standard for pointing to the integrity of data and the authenticity of the information presented? Any ideas?