View Full Version : Next Evolution of Search Engines - Internet - Computing
:eek:
The Futurist: The Intelligent Internet
http://www.gcn.com/vol1_no1/daily-updates/26338-1.html
The TechCast system, formerly called The GW Forecast, is a data-base-driven Web site in which panels of experts provide online estimates to carefully researched questions. The estimates are pooled automatically to produce the best possible forecast of when each technology is likely to take off, the associated confidence level, and size of the potential market - in real time.
* Internet search engines such as Google and Yahoo operate voice-recognition systems that help users find what they seek.
* Pattern matching and text parsing are used to improve searches by Google and AltaVista.
dannysullivan
06-24-2004, 07:48 AM
Google actually offers voice recognition now at its Google Voice Search (http://labs1.google.com/gvs.html) service. It got more press recently but has been in development there since 2001 (http://searchenginewatch.com/sereport/article.php/2163641) and made an official part of Google Labs back in 2002 (http://searchenginewatch.com/searchday/article.php/2159971). The futurists don't seem to know the past very well :)
Will the next evolution of Search Engines Attempt "concept searching"
again - did Excite.com demise as an unique algorithmic search engine - combined - with Google's Link Popularity success - mean the end of this experiment or goal. Will "synonym searching" be the best available option :confused:
http://google.blogspace.com/archives/001018
~~~~~~~~~~~~~~
Also, is it too computationally demanding to do it by today's competative standards - and today's vastly increasing amount of Web pages. :o
http://www.google.com/search?hl=en&lr=&ie=UTF-8&q=excite+concept+searching
(a blast from the past...)
mesadynamics
08-02-2004, 02:22 PM
Will the next evolution of Search Engines Attempt "concept searching" again
...
Also, is it too computationally demanding to do it by today's competative standards - and today's vastly increasing amount of Web pages.
The solution for doing concept searching is to move the computational demand to the client machine and let Google and other search engines simply act as a "feeder" of raw data for your search.
This, in fact, is exactly how my company's product works: by data mining the text from pages and pages of Google results, it is possible to uncover "concepts" that you would never find looking through the search results yourself.
If you're interested you can check out our site (http://www.mesadynamics.com/theconcept.htm).
I really think the computational effort is too much for the server side to handle, especially given the sheer number of requests, but who knows, maybe we'll see something like this offered as a service in the future...and put me out of work. :(