The Quest for the 3rd Generation Search Engine Continues!

A few months ago New York Times probed a bunch of start-up search engines as possible way to go in beating Google's dominance on the search market. These days yet another influential media giant Newsweek seems to be going against Google. In a quest for the best search engine Newsweek, in a story titled "Searching For The Best Engine", claims a global effort is underway to invent a better way of finding things on the Web and asks the rhetorical question "Could Google be vulnerable?" .

Aside the French, German and Japan's governments pouring hundreds of millions in an effort to establish local search companies as well as the traditional Google rivals in the face of Microsoft, Yahoo and somehow Ask.com, Newsweek seems to be focused on the so called 3rd generation or alternative search engines emerging upon us with innovative ideas, concepts and technologies promising to deliver better results than Google in one way or another.

In 2005 and 2006, venture-capital firms invested $350 million into 79 search-related start-ups. Knight, the author of AltSearchEngines.com's blog tracks no fewer than 1,000 search contenders, mostly U.S.-based, that have something to recommend the old guys in the sector.

The story mentions a number of innovative and highly promising start-up search engines, technologies and approaches as follows:

Hakia – Relies on natural language processing.

Wikia - The one the Wikipedia's founder stays behind.

Naver* - a South Korean search veteran, has been around for pretty much the last decade.

Quintura – A visual engine started out by Russians but now said to be based in Virginia.

NosyJoe – A Social search engine that employs intelligent tagging technologies and runs on a semantic platform.

Mahalo - Hard to say search engine it more looks like a directory with quality content handpicked by editors. Jason Calacanis is the one behind the company.

ChaCha – Real humans try to help you in your quest for information, via chat.

Powerset – Still in closed beta, well over-hyped and seems to be also relying on understanding the natural language.

Clusty* - founded in 2000 by three Carnegie Mellon University scientists.

Ask.com* - Today's #4 search engine immediately after Google, Yahoo and MS.

Lexxe - Sydney based engine featuring natural language processing technologies.

Accoona – What's interesting for the company is that they have recently filed for an IPO in US planning to raise $80M from the public.

KartOO – A French initiative entirely done in a flash, visual meta search engine.

KoolTorch – Seems to be yet another visual search engine, it is said to be also French one. Weak point could be that they don't use their own index and seem to be using the Dmoz.org's results.

Squidoo – It has been started in October 2005 by Seth Godin and looks more like a wiki site, ala Wikia or Wikipedia where anyone creates articles on different topics.

Sproose – Appears to be also using third party search feeds; ranks the results based on the users' votes.

Trulia – A real estate search engine

Healthline - Enables you to plug in symptoms and track down possible causes and treatment, health search engines, a heavily competitive area where Microsoft is also playing a role in, among others

Globaspec – Searches for industrial information and is aimed at engineers.

Like.com – Product search engine.

Spock – Focuses on people information, people search engine, and there are other competitors in the sector.

* Note that Naver is not a start-up search engine; it has been around for quite long time and is the lading search engine in South Korea . Ask.com and Clusty.com are also among the old guys from the sector.

Very interesting fact is that right now all these underdog search engines are said to have a combined market share of less than 5 percent of all queries, which is an impressive number knowing that 1% of the search market is today valued at the staggering amount of more than $1B, according Don Dodge.

The story is scheduled to appear on the Nov 5, 2007 print issue of Newsweek.

Just like a few months ago when we covered the NYT's story today we again think that only the time will tell us who will be the big winner at the end of the day. On the other hand whatever the case is any of the above search engines may get acquired sooner or later for anything in the 8 digit range (excluding some of the major players from the list above which may hunt amounts down in the 9+ digit range), which, in our feeling, is the main goal of most of these start-ups.

On the contrary, many IT experts and professionals do believe that undermining Google is most likely a task that will not be accomplished by a search engine in our traditional understanding, but rather would come from a concept that is anything else but not a search engine. Such concepts, including but not limited to, are Wikipedia, Del.icio.us and LinkedWords. In other words finding information on web doesn't necessarily mean to search for it.

Some more information on the topic:

www.newsweek.com/id/62254/page/1

altsearchengines.com/2007/10/05/the-top-10-stealth-search-engines/

altsearchengines.com/2007/10/30/why-is-altsearchengines-in-newsweek-magazine/

userweb.port.ac.uk/~duke-wie/blog/2007/10/30/searching-....

www.nytimes.com/2007/06/24/business/yourmoney/24digi.html

dondodge.typepad.com/the_next_big_thing...

Is NosyJoe.com the next Clusty.com?

Clusty.com is a well known search engine project basically a spin off from Vivisimo. Vivisimo on the other hand was founded in 2000 by three Carnegie Mellon University scientists who decided to tackle the problem of information overload in web search. Rather than focusing just on search engine result ranking, they thought that grouping results into topics, or "clustering," made for better search and discovery.

Clusty in general is a word derived from clustering, in our best guess, and the search engine itself became popular with its function to cluster the web, basically explained. Yet it is know that Clusty queries several top search engines which in its basic means a metasearch, i.e. no own index.

Initially planned to take on Google (Carnegie Mellon University vs. Stanford) Clusty seems did not make it at all and failed to gain the basic popularity necessary to rank among the first 10 search engines in terms of market share. The latest phrase seen in the public about Clusty is: “It (Vivisimo) also has a consumer search engine, Clusty, that allows Web browsers to query several search engines at once.”

On the other hand their clustering technology is considered pretty good one but seemingly other start ups are working on similar ideas, one of them, is NosyJoe -- a small start up search engine -- that is taking the content clustering to the next level.

NosyJoe is still in private beta, however, it was mentioned in NYT this last June and since then the project gained some traction among the search experts on web.

Basically NosyJoe extracts the meaningful sentences, phrases and keywords from the content submitted by the users. It then makes the information pretty much meaningfully searchable, web visible and findable and also clusters it into contextual channels and tags and applies a set of algorithms and user patterns to further ranking, mashing and sharing.

From our quick tests on the site the search results returned were presented in form of meaningful sentences and contextual phrasal tags (as an option), which turns their search results into -- something I have never seen on web so far -- neat content components, readable and easily understandable sentences as search results unlike what we are all used to, some excerpts from the content where the keyword is found in. When compared to other search engines’ results NosyJoe.com’s ones appear truly meaningful. You can check it out for yourself, we have submitted the about page of Clusty to NosyJoe and here is what the site shows it has for the Clusty.com’s domain: http://www.nosyjoe.com/website/www.clusty.com

Another search query done in NosyJoe for Clusty yielded pretty good meaningful results too:
http://www.nosyjoe.com/search.php?q=clusty

Another start up search engine known to be playing in the clustering arena is Quintura, a US based company started out by Russians.

So is NosyJoe.com really the next clustering engine or it is yet another social search experiment.