The latest entrant in the search game isn’t looking to capture another
vertical niche or tiny piece of the search pie. No, Powerset said it plans
to challenge market leader Google with a natural
language search engine.
The San Francisco-based company is licensing
technology from the famed Palo Alto Research Center (PARC), where the first computers to sport a graphical user interface and other breakthrough research has been done. PARC is a wholly owned subsidiary of Xerox.
Steve Newcomb, a co-founder and chief operating officer of Powerset, said
the company plans to offer users something quite different than search
offered by Google, Yahoo, Microsoft and others.
“Google built the first ‘Model T’ of search and they did an awesome job.
We love Google,” Newcomb told internetnews.com. “The next phase in
search is like when cars moved to an automatic transmission. The idea was to
make them function more like humans. Five or ten years from now we’ll look
back and say ‘Remember how we use to use keywords for search?'”
Powerset plans to make a beta of the service available later this year.
Newcomb said the public site will be able to understand the intent of a
query typed in a complete sentence. The company also plans multi-lingual
versions.
If that sounds like a difficult task, that’s putting it mildly.
Researchers at PARC have been working on the challenge of linguistic search
for over thirty years. Powerset’s founders, familiar with the effort, formed
the company 18 months ago with the idea to work with PARC to launch a new
search engine.
Newcomb said his company also has its own technology and will be announcing shortly it has hired some well-known researchers in the search field. Powerset also plans to announce alliances it has in the works with other companies. Amazon is one partner already announced.
Such deals will be necessary to provide the considerable resources needed
for a natural language query service able to tap the entire Internet – or at
least offer something comparable to what the other major search players do.
“We’re going to be processing a ton of data, more than what a keyword
search does, because we have to do more,” said Newcomb.
Danny Bobrow, a key senior researcher at PARC, said the technology behind
Powerset has the potential to provide a radically different search
experience. He explained that a lot of smaller research hurdles had to be
cleared to get the technology ready for prime time.
“Just one example, we did a lot of work related to the morphology of
words, so the system knows ‘was’ is past tense of ‘is’ and ‘growing’ comes
from ‘grow’ and also all the spelling issues.”
Bobrow noted that PARC’s relationship with Powerset, which includes
licensing of patents and revenue sharing, is unique in that PARC researchers
continue to develop the technology. “Most startups have to invent everything
themselves, but here we can continue to explore alternatives while
[Powerset] launches the business,” Bobrow told internetnews.com.
Newcomb said Powerset doesn’t plan to “reinvent the wheel” and will
likely use a revenue model similar to Google and others, i.e. an ad network
related to search results. Bobrow said that despite the complexity
underlying Powerset’s effort, results will be delivered to users in the same
sub-second speed users have come to expect.
“I don’t think there’s a product if people have to wait for many
seconds,” he said.
Bobrow gave a consumer example of how the Powerset service works. When someone types in “Who was Spielberg married to before Kate Capshaw” Google and others give results related
to the movie director Steven Spielberg and actress Kate Capshaw.
“Google doesn’t give you the answer, Amy Irving, because it’s not part of the
question. What you really want is the answer, not hundreds or thousands of
links. We give you the answer.”
There are already other search competitors using alternative methods to compete with Google. One is Hakia, which uses a ranking algorithm it calls SemanticRank. Hakia is in beta and publicly available.
“There are going to be a wave of natural language companies coming through,” said Newcomb. “We think our approach is better, but we’re friends with those guys and we want natural language to succeed.”
Unlike Google, Hakia produced the correct, if a bit folksy, answer to the Spielberg query. The top of its results page said: “You are very curious today. Spielberg was once married to Amy Irving and is now married to Kate Capshaw.” That was followed by links to pages related to Spielberg.
