As Tim Berners-Lee is the first to acknowledge, today's World Wide Web can be a difficult place to get things done. Your search engine can't tell the difference between a Zip Code and phone number. To a computer perusing a travel site, a departure time of "09:05A" could just as easily mean 09.05 Australian dollars. On the Semantic Web, words will be tagged in a language called XML so computers can tell what they mean. And smart software programs called "agents" will be able to grasp both the meaning and context.
1. JANE ASKS HER AGENT FOR HELP
Jane is planning a vacation in Fiji, and--thanks to the Semantic Web--her computer will make all the necessary arrangements. Her travel agent, built entirely from software code, will explore the Web, gather all the relevant information about airline schedules, condominium prices, and the merits of the different beaches, and return with Jane's digital tickets in hand.
2. THE AGENT EXPLORES THE SEMANTIC WEB
The agent's first stop is a virtual library, where he learns the terms and definitions used by airlines, ships, real estate agents, and car-rental services. Each industry uses its own definitions, written in a language called XML. Once he knows the "tags" for concepts such as , he can find the info Jane needs.
3. GOING TO MEET THE AIRLINE REP
Armed with his semantic data, Jane's agent is now ready to negotiate with the software reps on various airline Web sites. In addition to learning about fares and flight schedules, he picks up tidbits about alternative travel modes, such as cruise ships.
4. THE RIGHT PLACE TO STAY
After boning up on the various travel alternatives, the agent heads back to the library to read up on the semantics that describe condo rentals. Then he zips off to the Web sites of real estate agents handling properties on Fiji, including resort chains and time-shares. Because Jane's request included the words "house" and "condo," he knows to ignore any hotel offerings that don't include kitchens.
5. NOW TO CUT THE OPTIONS
What emerges are a dozen candidate flights and scores of lodging alternatives. These can be juggled to yield thousands of combinations. Jane doesn't have the patience or time to explore them all--but the agent does. In a wink, he ranks all the possibilities based on priorities, such as maximum time spent on the beach. He picks five alternatives to show Jane.
6. JANE MAKES THE FINAL DECISION
She prefers air travel to cruises, so the agent dives back onto the Web to close the deal. He confirms some reservations, cancels others, and pays the software agent at United Airlines. He concludes the deal for a condo and reserves a car for Jane--without waiting for her to ask--knowing she usually rents sporty two-door cars.