Tim Berners-Lee must feel like he’s in a time warp. In the early 1990s, he spent a frustrating year trying to get people to grasp the power and beauty of his idea for a scheme known as an Internet hypertext system, to which he gave the beguiling name the World Wide Web. But since the Web didn’t yet exist, most people couldn’t imagine the implications of what he was talking about. Berners-Lee persevered, and with the help of the few people who shared his vision, his invention became the fastest-growing media distribution system in history.
A decade later, Berners-Lee is struggling with the same problem-only this time, he’s trying to articulate his dream of a Semantic Web. The idea is to weave a Web that not only links documents to each other but also recognizes the meaning of the information in those documents-a task that people can ordinarily do quite well but is a tall order for computers, which can’t tell if “head” means the leader of an organization or the thing on top of a body. “The Semantic Web is really data that is processable by machine,” says Berners-Lee, who is director of the MIT-based World Wide Web Consortium. “That’s what the fuss is about.”
Today’s World Wide Web is fundamentally a publishing medium-a place to store and share images and text. Adding semantics will radically change the nature of the Web-from a place where information is merely displayed to one where it is interpreted, exchanged and processed. Semantic-enabled search agents will be able to collect machine-readable data from diverse sources, process it and infer new facts. Programs that weren’t made to be compatible with each other will share previously unmixable data. In other words, the ultimate goal of the Semantic Web is to give users near omniscience over the vast resources of the Internet, turning the millions of existing database islands into a single gigantic database Pangea.
To compare the Semantic Web with today’s Web, Berners-Lee-an intense person who speaks in low-volume bursts-offers the following scenario: Imagine registering for a conference online.
The conference Web site lists the event time, date and location, along with information about the nearest airport and a hotel that offers attendees a discount. With today’s Web, you have to first check to make sure your schedule is clear, and if it is you have to cut and paste the time and date into your calendar program. Then you need to make flight and hotel arrangements, either by calling reservations desks, or by going to their Web sites.
“There’s no way you can just say, I want to go to that event,’” explains Berners-Lee, “because the semantics of which bit is the date and which bit is the time has been lost.” But on the Semantic Web, he asserts, those bits will be labeled; the software on your computer will recognize those labels and automatically book your flight to the conference and reserve a hotel room with the click of a button.
The Semantic Web will also be a richer, more customizable Web. Imagine running your cursor over the name of the hotel and being informed that 15 percent of the people who’ve voted on its quality say it’s excellent. If you happen to know that the hotel is a dump, you can instruct your browser to assign those people a trust level of zero. (The polling information would be saved on a third-party “annotation server” that your Web browser accessed automatically.) By assigning high levels of trust to people who match your tastes and interests, and “bozo-filtering” the people who don’t, the Web will start looking more like your Web.