Predicting the Future of the Web: The Abstraction Vector

I recently posted on Edgeio, an interesting new service that parses listings of items for sale out of appropriately-tagged blog posts in The Long Tail.  Then I speculated about other types of information that might lend themselves to similar services, like events. 

When software engineers foresee that they will have to do the same thing several times, they try to automate the process they would otherwise have to do manually.  In order to accomodate variations in the tasks, they create "abstractions" that generalize elements of these variations that are held in common.

Once more via TechCrunch (which has become the de facto chronicler of the Web 2.0 landscape), I learned today about Vast via Nik Cubrilovic's post.  Vast does what Edgeio does, but across multiple different kinds of information, including listings, but also user profiles for example.

It will be interesting to see how the search engines respond to this.  Parsing the web and aggregating popular kinds of information in semi-structured data formats seems like a very important capability, one that leads us from "search" through "parametric search" and perhaps ultimately to SQL-like queries of The Giant Database In The Sky.   It's interesting to think about the business possibilities this suggests.