CAMBRIDGE -- The inventor of the World Wide Web told a technology conference yesterday that making the Web more useful hinges on a familiar challenge: getting the players behind the technology to agree on standards governing how computers communicate with one another.
That obstacle, which confronted the initial development of the Web, looms large again in the nascent stages of what Tim Berners-Lee calls the ''semantic Web," an evolutionary process to make more kinds of data easier for computers to locate and process.
''It's all about standards," Berners-Lee told an audience of about 500 in a speech opening Technology Review magazine's two-day Emerging Technologies Conference at the Massachusetts Institute of Technology. ''Standards are the basis for any emergent technology."
The obstacles in the way of advancing Web development are as much social as they are technological, and the industry must avoid the temptation to lock up key technologies by demanding royalty payments, he said.
Berners-Lee, a 49-year-old native of England who runs the standard-setting World Wide Web Consortium from an office at MIT, envisions a new phase of the Web in which the various sources of information can more readily interact with one another.
Rather than merely navigating their way via Web links to information related to their interests, Web surfers should be able to manipulate it to intelligently steer them to data with specific meaning to that person, he said.
Berners-Lee cited the example of a Web ad for a seminar. While the computer user may know what the information means, the computer doesn't. Someone planning to attend would have to note the date by pasting data into an electronic calendar, or add the names of people taking part in the seminar into an address book.
Berners-Lee envisions encoding the information in a way that enables the computer to comprehend the data and link it to applications, automatically adding information about the seminar into a calendar or address book. That involves standardizing how information is stored on the Internet. Web data would carry tags to give them meaning, so computers can do a better job of searching.
Ultimately, he said, software could be written to process the data and make inferences that previously required human intervention.