Web::Vast - a new mashup opportunity
Techcrunch profiled a new service called vast today. Here is an excerpt from their FAQ page:
Vast is a search service - we scour the web across millions of sites and billions of pages to find nuggets of information - jobs, cars, people, and other classifieds, and offer the resulting database to end-users and other web sites. We currently offer the largest database of cars and jobs, and one of the largest databases of people profiles on the entire Internet.
The Vast crawler even digs deep to find some of these listings, which you won't find currently on any search engine! The Vast API is available, free and unlimited, for commercial and non-commercial use. Vast does not act as a broker, agent, or dealer, and the results aggregated by Vast always point back to the original listing for all of the information and to complete the transaction. Think of it as Froogle for classifieds, but with an open interface for partner sites and developers.
This is very exciting for me for several reasons:
- I'm interested in any new services related to classifieds or personals
- I am fascinated with any search / crawl technologies that 'interprets' the content for classification
- They are letting other websites to use the indices for free
They crawl over 3 billion pages, so that will include the core and the edge of the web. Then they filter out irrelevant pages using their own unique technology. As a result, you get only relevant items from any page on the web.
I immediately thought this could be a potential competitor to edgeio, but Keith says that he believes its more of a competitor to oodle in the sense that edgeio users are publishing (submitting) their ads via edgeio, whereas oodle collects ads from the web. I think that's a great distinction point.
This is what Naval Ravikant says on this issue:
Not at all. EdgeIO focuses on blogs and covers a lot of categories. It has a few thousand listings. EdgeIO seems to be building a destination site as well, with some RSS exports.
Vast crawls the whole web and at the moment only covers three categories. Vast has about 16 Million listings. Vast is not meant to be a destination site but more of a content router for other sites.
EdgeIO has a good team and a powerful vision of the future, but the two sites are pretty different.
I like vast especially, and most importantly because they are letting other web sites to use their database. Someone just asked (in the comment section at techcrunch) how this is different from Kosmix, and I believe this openness is the biggest difference. I guess they have different vision and philosophy. On the bottom of their website, there is a link titled "Steal This Site", which leads you to their API page. I can imagine tons of new applications that creatively utilize (social bookmarking, rating, tagging, etc.) their data popping up in the near future. This is great for small companies and great for the consumers!



















I noted in my comment on TechCruch that I think the difference is between finding and normalizing structured data and categorizing the Web. There are some simliarities, but our technologies are quite different and sites will look quite different in a year.
Thanks for cheerleading web classification! =)
Posted by: Mark Johnson | March 15, 2006 at 02:56 PM