note on company name - it's Novo|seek with a vertical bar between the words & seek is in italics

Not the "next Google" but still interesting

By Scott Bradner

No startup currently on the horizon is likely to overcome Google's breadth of services and Internet mindshare but there are some recent developments that may just pull users away from Google -- at least for some functions.

There is no question about Google's dominance in Internet services of all kinds. They are, by far, number one when it comes to search engines -- 63% of the US searches in February 2009 according to comScore, with Yahoo! a distant second with 21% followed by Microsoft, Ask and AOL in single digits. In addition, Google's wide array of other products such as gmail, Google Docs, Google Analytics, Google Health and, most recently, Google Voice have caused Google to become synonymous with Internet in many people's minds. But, even with Google's overall Internet dominance and its myriad of specialty search functions, other companies are still looking for business opportunities. The companies offering specialty search engines are the best example of this activity.

Shortly after my column on double paying for the results of federally funded research (Paying for Internet content once is not enough - http://www.networkworld.com/columnists/2009/031109bradner.html) was posted I heard from a spokesperson for Novo|seek (http://novoseek.com/). It turns out that Novo|seek is a user of the type of research information I wrote about. The spokesperson arranged for me to speak with Ramon Alonso-Allende director of marketing and business development for Madrid, Span based Novo|seek. Novo|seek is a specialty natural language-based search engine that focuses on biomedical literature. The search engine does not just do simple word searches, instead it understands how terms are used in the literature so it can find documents where synonyms are used and ignore documents where a search term is used with a different meaning.

Novo|seek is, by far, not the only specialty search engine on the Internet, although it seems to be one of the ones best suited to its intended users. A few days ago the New York Times had an article on Kosmix, a startup that tries to return "more about something" than simple word-based search engines do. A few days before they published the article on Kosmix the Times wrote up Anglesoft, a search engine targeted at entrepreneurs looking for investors and OneRiot, a search engine that looks for gossip.

A quick web search, done with Google naturally, found many other examples including

o PeoplePond (http://www.peoplepond.com/), optimized for searching out people,

o Fashion Latte, a "visual search engine" developed by a bunch of University of Illinois students to facilitate people browsing dresses,

o dogpile (http://www.dogpile.com/), which looks like a normal search engine but some of the revenue it generates goes to the American Society for the Prevention of Cruelty to Animals,

o the EFF FOIA Search Engine (https://www.eff.org/issues/foia/search) from the Electronic Frontier Foundation and designed to search documents released under the US Freedom Of Information Act.

o FanSnap (http://www.fansnap.com/), for finding concert, sports and theatre tickets

o Hunch http://www.hunch.com/, "a consumer-facing online decision engine' that just got $2 M in venture funding and is supposed to go live soon

o Wolfram Alpha (http://www.wolframalpha.com/), a natural language based search engine that is supposed to go live in two months

Natural language-based search engines seems to be the buzz of the moment, seemingly forgetting that this was one of the original concepts behind Ask Jevves (now ask.com (http://www.ask.com/) and, for their rivals, betting that Google somehow can't figure this technology out -- not a good bet to me.

disclaimer: The math of betting has been the subject of Harvard courses, and not-quite betting was the subject of the movie 21 that featured a Harvard character but the above is my own review not Harvard's.