Now, It's The "Vectories" That Are Coming!
Back in April, I wrote an article about an expected growth in vertical search engines or "vortals," as they are sometimes called. This was due to the emergence of new services that I feel make it cheaper and easier for sites to create their own crawler-based search engines that focus on particular topics. Now another group of new services are emerging, those that offer to automatically categorize web pages into Yahoo-like directories. For example, point these services at a set of pages from different sites about cars, and they'll attempt to automatically group the pages into categories that make sense. As a result, we may see more vertical directories, or "vectories," as I'll dub them, also appearing across the web in the near future.
In general, the vectory products aren't magic. You typically have to provide "training documents" for the major categories you wish to populate. You give the software some good examples of pages about cars and trucks, and it will then use them to group other pages into either a cars or trucks category. Human editors may also be involved to create keyword definitions for some categories and to filter out obviously incorrect matches.
Here's a look at some of the players in the vectory market:
Mohomine is aimed at both the vertical portal market and those needing Intranet search solutions. It can create vectories for consumers or organize a company's internal documents into categories. Mohomine also has document discovery capabilities. Show it examples of documents you like, and it can mine the entire web, if you desire, to recover similar matches. Mohomine can classify more than web pages. It can extract textual information from objects such as computer code, allowing such items to be classified alongside HTML documents. The company launched services in March.
Mohomine-powered search engine that helps software developers track down source code.
Twirlix offers both a crawler-based search engine and an automated directory to its clients. Sites can be ranked by using criteria such as link analysis, the size of a site and freshness of content to determine relevancy. Thumbnail images of each page can also be presented. The company launched its US services in May and has been open in the German market since last November (http://www.twirlix.de/). The Twirlix US and German home pages serve both as a technology demo to potential clients and as a web search site for consumers.
Autodesk vertical portal designed for designers and engineers. Twirlix recently signed an agreement to provide services for PointA, but I don't believe any new material has yet gone live.
Founded by two students from Carnegie Mellon University, the birthplace of Lycos, Hubat offers automated directory building. The site serves as a demonstration of the company's technology and as a resource for consumers to search the web.
Hubat: A promising automatic directory builder
TBTF, Feb. 7, 2000
Tasty Bits From The Technology Front author Keith Dawson came away impressed with Hubat's automatically generated page descriptions, in this review.
Released in April, the SeeAll tool is designed to create directories within an Intranet enviroment. It can also be applied against web-wide documents.
RuleSpace offers the ability to categorize documents for purposes of searching. However, the product really seems positioned as a classification tool for networks. For example, you might use RuleSpace to identify and prevent porn documents from being accessed on corporate networks or to filter out certain types of documents from a directory. The company has been operating since 1996. It lists AOL and Amazon's Alexa as companies using its technology to help categorize information.
Inktomi Directory Engine
Inktomi has offered its "Directory Engine" since July 1999. The product that automatically classifies web sites into categories and makes use of link and caching data to help determine which documents are most popular. The company also recently acquired Ultraseek, which has categorization technology designed for use on Intranets.
The Vortals Are Coming!
The Search Engine Report, April 4, 2000
The earlier article I wrote about vertical search engines, with players mentioned.
Vortal Services Now Offered
The Search Engine Report, July 5, 2000
Short follow-up to the vortal article.
Meet Your Favorite Search Engine Watch Contributors
Many of SEW's leading expert contributors will be at ClickZ Live, the new online and digital marketing event kicking off in New York (March 31-April 3). Hear from the likes of: Thom Craver, Josh Braaten, Lisa Barone, Simon Heseltine, Josh McCoy, Lisa Raehsler, Greg Jarboe, Dan Cristo, Joseph Kerschbaum, John Gagnon, Eric Enge and more!