Let’s start with a high-level view of Deep Web Technologies…
Deep Web Technologies started in 2002. We’re about 20 people, based in Santa Fe, New Mexico, and we started out by building public federated search applications for federal government clients. Our best known application is probably Science.gov. It’s a public site and searches about 50 different R&D data bases from across the U.S. federal government. We followed on with worldwidescience.org that searches about 80+ databases from all over the world, including databases that are not in English. We have a very unique patent-pending technology that does multi-lingual searching, so we can take a query, translate it from the user’s language to the languages of the databases being searched, which include German, French, Russian, Chinese — about 10 languages total.
And then from there, we expanded into working with commercial customers. One of our customers is BASF [major German chemical company]. We’ve also expanded into providing federated search to academic libraries. Numerically, they’re the largest number of our customers, and the marquee customer on that list is Stanford University.
The model for the company is mostly subscription based. Our customers, our partners pay us a subscription to develop, run and maintain applications. Stanford University, for example, pays us an annual fee to run the application for them. In the future, ad-based revenue can be a viable part of what we do. The mobile apps — Biznar and Mednar — are free.
Tell me about your Biznar mobile app — where does it fit in your larger mobile strategy?
So far, Biznar and Mednar are about us ‘dipping our toes into the water’ in mobile applications. Mobile isn’t a source of revenue, and the apps are free, so we have released our mobile apps without much promotion to watch the user response first. Having said that I think what we offer works well on an iPhone for people who are heading to a business meeting and need one last fact or piece of information to prepare for that meeting. Now, we get asked all of the time: “How is Biznar — or any of our apps — different from Google?” The difference is that we search quality information sources. Google has great stuff in its results, but it also has a lot of sources that are not so reputable or so useful. You don’t have that problem with Biznar.
Today we’re searching 70 sites. Because we’re doing a real-time search we’re getting information as it’s available on the sites we’re searching. Charles [Knight] points out that we’re also always listening to input from users about sites and sources they would like us to include. That’s great and, as a small company, we can be pretty responsive to their recommendations.
You’re all about business search, a focus and a capability that you could also extend to other companies that want to built services or apps…
I’m always looking at opportunities through the eyes of federated search. So, I’m not so interested in creating unique content, but I am interested in providing related content. We’re working with a number of partners now who will each have their own niche, and we’re looking at providing related content. One company we’re working with is Data Planet . They provide access to all kinds of statistical data from government agencies and the census bureau, as well as some commercial sources. What we are doing is building an application that searches their data and will also provide context through related content to the data that they provide. Say you’re looking for a chart on teenage pregnancies in Montgomery County. We could provide some related content to that search like a news story about teenage pregnancies in general, or other information specific to Montgomery County. I think there’s a lot of opportunity in providing content that complements what someone is looking at or searching for.
My bigger vision here: to create a catalog of thousands of information sources and make it easy for a user to pick their own sources that they want to search — either the ones they’re familiar with, or sources others have recommended. Imagine a user finds a source that really meets their needs and they could go to the catalog and ask ‘who are the other people who use this?’ Similar to the Amazon model. So, what other people use this source, or search this source? What else are they searching for? And my even bigger vision: translate that into mobile and smart devices to make it easy for any user to basically create their killer app.
Hmm.. sounds like you could imagine becoming a platform tool provider allowing me — or an app developer — to use your approach to federated search and catalogs of sources to create my search engines on top of your technology. Is that correct?
Absolutely. We definitely want to move much more into becoming the platform that allows partners and customers to be able to configure what they’re providing to their users. Take health-related information. Imagine that in our catalog we have a selection of sites that are hepatitis related for example. Some health organization, or some individual that’s created a health related site, can leverage access to related content tailored to their needs that can be provided through our apps. A lot of this is in the roadmap, but we have the ability now. We also have an alerts capability that can be set up to return results on a daily basis, weekly basis of new information that meets a particular user profile.
Finally, we know Charles is sharply focused on iOS, and covers your iPhone app in his own road test. What are your plans for other platforms?
I think it’s important to get an Android app working. I don’t have a specific timeframe for that. There are also engineers in my company who think we should also be looking at HTML5. So, we’ll have to consider that as well.
- Deep Web Technologies: Cracking Multilingual Search 1 (arnoldit.com)
- A greater need than ever for multilingual federated search (deepwebtechblog.com)
- THE Future of Mobile; Authorities Weigh In On Mobile Megatrends (mobilegroove.com)
- Mobile Search App Review & Road Test: Biznar (mobilegroove.com)