The Rise of Google

In April 2006, Google purchased a new search engine algorithm, Orion, from Ori Allon, a twenty-six-year-old Ph.D. candidate at the University of New South Wales, in Australia. They also hired Allon. Historically, Google has not revamped its engine so that it hasn't seemed unfamiliar, but rather it incorporates new features or technologies. The key benefits of Orion seem to be that it can gather results not only by keyword but also by words that are associated to the keyword, a function with which all search engines have struggled. It's not yet clear whether the engine, as improved by Allon, will know that a search for a sports car should send you to an MG, but it will, to use his example, take a query about the American Revolution and return results about American history, as well as George Washington and the Declaration of Independence. Orion's other advantage is that it can display excerpts of the relevant pages without the user having to leave the search engine site.

Another approach search engines are taking is clustering. Mooter and Vivísimo, for example, are two search engines that will take a query about “bass” and immediately break it into clusters of related information. What that means is that the results for the fish called bass will be separated from those concerning the musical instrument, the beer brand name, or the shoe manufacturer. Given these options, the person conducting the search can start in the right ballpark, at least, immediately eliminating Web sites that are totally irrelevant. Mooter will also personalize rankings as the user makes choices, using artificial intelligence to interpret feedback from user behavior.

The search engine Teoma (the Gaelic word for “expert”) uses authorities on subjects to rank results and identify key authoritative pages. These have been established by looking at subject-specific communities. Paul Gardi, former president and CEO of Teoma, explained in 2004, “When you get down to the local level [which he says Google does not], you will find that links cluster around certain subjects or themes, very much like communities.” Future search engines will undoubtedly utilize experts to sort information or they will instead be programmed using artificial intelligence and frequent user preferences.