Sent an exploratory note to DeepWeb Technologies the other day, asking if they’d tell me a little more about their product and give me some idea of what it might cost to use their Explorit software to build a few federated search systems. If you’re not aware of DeepWeb Technologies and their software, let me refer you to science.gov (a nice showcase for the Explorit package).
This afternoon I got a call from a Deep Web rep, asking if I’d give him some idea of how I envisioned using their software. “Sure,” I said, “we’re building a number tightly-focused research portals for our academic programs here at the university and it occurred to me that building small, tightly-focused federated search engines that worked against a small cluster of relevant sources in each discipline might be a nice service to add.”
I expected a response like, “Well, our product is focused on bringing order to the internet and your proposed application is a bit small for our product” but I was wrong. Instead, he said, “Well, you’ve just described the exact project we’re working on with Stanford University and we hope to release it as a product early next year.” Now I’m really interested.
After the call I did a quick check of Stanford’s web but most of what they appear to be working on is hidden behind authentication screens. I did find one open example:
https://deepweb.stanford.edu/digcolls/search.html
To see another example of DeepWeb’s work, let me recommend you take a look at Biznar, a federated search engine for business sources.  Information Today posted a review of Biznar a few weeks ago that does a good job explaining the system.
While on the subject, let me point out that DeepWeb Technologies sponsors a blog that’s certainly worth following if you have any interest in the topic of federated searching:
http://federatedsearchblog.com