UBC spinoff plays on ability to absorb tastes of individuals
Gillian Shaw
Sun
Mike Klaas (left) and CEO Ali Davar discuss the UBC spinoff that came out of the university’s laboratory for computational intelligence. They have taken online search engines to the next generation with Worio, a personalized service. Photograph by: Mark van Manen, Vancouver Sun
A University of B.C. spinoff is building on current search-engine technology to deliver the next generation of tailored search offerings that could only come from a service that gets to know you almost as well as your mother.
Worio, technology that came out of the university’s laboratory for computational intelligence, is on the leading edge of a trend in which personalization and contextual search will replace the more scattergun approach that can come with simple keyword searches and other online interaction.
“We are developing what we call a web scale discovery engine,” said Ali Davar, chief executive officer of the now-Yaletown-based startup. It received Precarn (technology development) and National Research Council funding, along with $3 million in angel funding to develop the technology.
Davar said Worio provides discoveries — that is, added information delivered alongside search results — based on context and not just keywords.
“Discovery is different than search, the purpose is to understand automatically what you are interested in already and push you that information.”
Walking through examples of Worio in operation, Davar pointed out that when he looked up ‘semantic web,’ Worio not only returned the results you might expect from Google or other search engines, but added information he might not have known to look up.
“I was pushed a recommendation for the Vancouver semantic web meetup group,” he said. “Worio was able to understand this was something I might like.
“If someone had told me about this group I could go into Google and put in those words but the simple fact of the matter was that I didn’t know to go looking for it in the first place.”
While there are examples, such as Pandora music service that offer similar service but on a narrower topic range, Davar said the approach doesn’t work when trying to generalize to the entire web.
“We don’t have as much of a controlled environment as those smaller domains,” he said. “We have overcome those problems by using a combination of technologies.”
Among those is an automatic tagging system that considers a small slice of information that has been commented on or shared by users and generalizes that for the web.
“It basically brings order to the web by bringing a topical layer to it,” said Davar.
On the left-hand side of a user’s Worio home page are the keyword search results that would be returned by search engines such as Google or Bing.
The right-hand side features what Worio provides: Context-based discovery.
“We are trying to bring discovery to search,” said Davar.
And in social media meets search, Worio users can share information with friends and contacts, building a network and getting information relevant to their interests that may have been turned up in friends’ Worio research.
Users can tag information they find and share it with other users or they can choose to use the ‘private mode’ in which Worio doesn’t track preferences.
“It is somewhere between a traditional search engine and a stand-alone recommender system,” said Davar.
“It is bringing discovery into people’s day-to-day experience rather than making it a limited view of what the most popular things are that people have dug today.”
Early adopters are already trying out Worio, which is in beta, and a browser plug-in expected in the next two or three months.
© Copyright (c) The Vancouver Sun
