Know what I mean, again

As a followup to my last entry on search, MIT Technology Review discusses a few more search algorithms currently being sported by new entrants in the search engine battles.

Teoma: Similar pageranking system to Google, but concentrates on links highlighted by related “community” sites as being the “authority” on the topic at hand.

Mooter: Displays groups or “clusters” of topics related to the user’s search request. That way, the user can drill down to more specific results. User modeling is used to personalize and calibrate future search requests.

Dipsie: Utilizes a more advanced spider that can navigate forms and database interfaces, giving it the ability to crawl documents otherwise inaccessible.

AskMSR: To bring more eyeballs to MSN, MS needs better search. AskMSR uses fuzzy logic to answer natural language questions. It searches the web to give a shortlist of hypothetical answers. (It reminds me of the “ask the audience” lifeline in “Who Wants To Be A Millionnaire”)

Stuff I’ve Seen: A tried and true (if monopolistic) MS tactic: integrate search into Windows. Will display all items off your hard drive and the Internet (Office documents, appointments, addressbook contacts, webpages) related to the current topic you are working on (replying to email, drafting a presentation, writing an essay).

John Battelle offers some remarks.

Silverstein has a pretty good point – the cost of switching from one search to another is nil. I started off liking WebCrawler, shortly by AltaVista (when DEC was still around! Time flies.). I switched to Infoseek in 1998 shortly after AltaVista banned all Geocities sites. I was an early adopter of Google, and am still using it to this day. With no barriers to entry, I’m guessing things will be very different in five years.