Webware

Ask.com's Tomasz Imielinski discusses semantic search as Microsoft's Scott Prevost and Google's Peter Norvig look on.

(Credit: Tom Krazit/CNET News)

SAN JOSE, Calif.--If those chasing Google have anything to say about it, search on the Internet is going to become more about a conversation than an exchange of keywords.

Panelists from the four major search engines--Google, Yahoo, Bing, and Ask.com--joined Web search start-ups TrueKnowledge and Hakia at the Semantic Technology Conference to discuss the rise of semantic technology as the engine behind the still nascent Internet search industry. Semantic search, or the idea of divining a user's true intent from how they enter their queries and how Web data is structured, is an unfamiliar concept to the majority of Web surfers who tend to think Internet search is actually pretty good as it is.

It's not, according to Tomasz Imielinski, executive vice president, global search and answers at Ask.com. "Most users don't know how good search can be," he said, drawing an analogy to those who were satisfied with their portable music options until the iPod came along.

The W3C is devoting an entire week to the concept of semantic technology, which involves Web publishers and search engines working together to structure data in a way that can be presented in a more appealing way than the "ten blue links"--a dirty term in the search industry these days--with which most searchers have grown familiar.

Yahoo has been banging this drum for a few years, introducing products like Search Monkey to help Web publishers start organizing their content around semantic standards, said Andrew Tompkins, chief scientist at Yahoo Search. "Today on any major search engine, you'll see structured information about a restaurant," he said, basic things like phone numbers, address, or maybe a link to a map of its location. All of those things require agreement on standards to make it happen.

But semantic search is also about improving the ability of search engines to analyze the meaning of plain text on a page, said Scott Prevost, general manager and director of product at Microsoft's Powerset division. A search engine that knows how to take a query and produce exactly what a person is looking for on the first page of results will prove attractive over time, he said.

The goal of all this work is to make search more intuitive, more like asking a friend or colleague a question, said Riza Berkan, CEO of semantic start-up Hakia. "We believe search is going to move to more conversational techniques," he said.

That's music to Ask.com's ears, of course. The company announced Wednesday that it now has 300 million question and answer pairs in its database that Imielinkski thinks provide context around searches.

But none of this work on semantic technology has done anything to dislodge Google from its position atop the search world, which actually grew a bit stronger over the past month according to ComScore. Google's Peter Norvig acknowledged the benefits of semantic technology and agreed that Yahoo deserves credit for pushing semantic technology along. He drew applause from the several hundred attendees at the panel discussion when he discussed Google's decision to support RDFa semantic standards, announced last month at Searchology.

Still, there's an economic component to this debate that Google isn't quite buying. None of the panelists brought this up Wednesday, but last year Microsoft's Prevost admitted that the desire to make an end-run around Google's dominance of keyword-based search advertising is what has driven semantic technology research, at least to a certain degree. "If people aren't bidding on keywords, and are bidding on concepts, it could completely change the ball game," he said last August at the Search Engine Strategies conference.

To that end, Norvig argued Wednesday that the idea of conversational search is good for people who aren't quite sure what they are looking for, or who don't quite understand a certain topic. But those who do grasp a topic and want a fast answer are much more likely to use keyword searches, he said.

Corrected at 3:49 p.m.: This post originally misstated the title of Ask.com's Tomasz Imielinski. He is executive vice president, global search and answers. Corrected on Friday, 11:35 a.m., clarifying the W3C did not sponsor the conference.

PALO ALTO, Calif.--Though the favorite metaphor to describe the Web has long been a highway, or for some, a "series of tubes," the man credited with inventing it all thinks of the Web more like the human mind.

"Lots of people are doing research around the Web...and there are interesting results, but a lack of a core curriculum in the universities," Tim Berners-Lee told a gathering of scientists at HP Labs and other Silicon Valley executives here. "I've been told the Web has 10 to the 10 to the 11 (number of) Web sites. The brain we study as a complex system." So why not the Web?

What millions of Internet users take for granted every day--using the Web as a means to download movies, read the news, or check Facebook--will look drastically different five years from now, and that calls for study of it as a science, according to Berners-Lee and his colleagues at the Web Science Research Initiative . Launched a year ago, WSRI is a partnership between the Massachusetts Institute of Technology and University of Southampton in England, and is encouraging the study of both the social and technological implications of wide-scale use of the Web.

Tim Berners-Lee and Wendy Hall of WSRI visited HP Labs in Palo Alto.

(Credit: Erica Ogg/CNET News.com)

On a tour to encourage the adoption of Web science as a course of study at local universities, Berners-Lee spoke about what kind of challenges the increasingly social Web presents. Corralling the information about us out on the Web, identifying where it came from and who is allowed access to it are major issues that come up every day. Facebook's decision to combine user profiles with advertising is just one example.

But there are even more serious implications of a Web that is a growing collection of our personal information. Who owns it? And how do we determine how our information is used?

One example Berners-Lee gave is hospital records. It's still unclear how to be sure that doctors can have access to patient information to identify and treat you, but at the same time keep that information hidden from, say, your employer. There is no answer yet. "It's about building systems and understanding where data is coming from," he said. And though that will take time to come up with a new way of storing and organizing information on the Web, he and others are already working on it.

Phishing scams, spam, an overload of our current Web infrastructure, as well as the democracy of online communities, are each major ideas that need to be looked at with an academic eye, said Berners-Lee, rather than from a closed, proprietary, or corporate perspective. Berners-Lee has long advocated a universal and open Internet, and is one of the founders of the World Wide Web Consortium, the organization that supports open Web standards.

Though much of the future of the Web is wide open, one thing that will happen is that we won't be inputting our personal information into separate social networks, he said. In other words, we'll have one profile that compiles all information related to us and our social networks. "Right now, so many people are complaining that they have told one Web site who their friends are, and another one who their friends are...In five years time, I hope people will be programming not at the document level, but at the application level," he said. "You will have something which is an application which is consistent for looking at different aspects of people. It (will use) your role as their friend for putting together a very powerful, all-encompassing view of them (online)."