Rensselaer Magazine
Feature Articles At Rensselaer President's View Reader Mail Staying Connected Alumni News One Last Thing

“Ultimately we would like the Semantic Web to be invisible to users,” says McGuinness. “You never have to think about the fact that you turn a light switch on and electricity flows. Similarly you don’t want to have to think about the Semantic Web, but its intelligent agents will just work for you, and do background tasks for you. You give them operating instructions, and they act on your behalf.”

Constructing the Semantic Web is only one of a number of projects the Tetherless World constellation has on the drawing board. It also plans collaborations with a number of interdisciplinary initiatives that can apply the principles of the Semantic Web to specific research projects. The group also aims to develop the field of “Web science,” studying the ever-evolving structure and use of the Web. As the constellation name implies, the group also has a vision of a smarter Web leading to a more efficient flow of information that will improve the capabilities of mobile and wireless devices.

“The Semantic Web will enable the less tethered world that we all can live in,” says McGuinness. “We’re going to have a much more tetherless society in the future, where you’re not required to be next to a large computer to do intelligent tasks. Today, you can kind of do some of that with your cellphone and PDA, from hotspots. But the best you can do is send marginally intelligent bits.” A smarter Web means smaller mobile devices could zero in on the tasks desktop computers handle today.

“In the future, the Semantic Web will place fewer demands on you to be physically connected,” McGuinness says. “What we’re really trying to do is give you more free time.”The constellation has lofty goals. “A very important part of our constellation goal is to change the world,” says Hendler.

“Will we achieve that or not? Only history will tell.” The man knighted by Queen Elizabeth II for his revolutionary creation of the Web believes this research team is in a good position to make that next leap forward. “For RPI to get Jim and Deborah together in the same place is quite a coup,” says Berners-Lee, now a senior researcher at MIT and director of the World Wide Web Consortium, which sets standards for the network. “It really makes Rensselaer a center of excellence in the field.


The Way It Works.

A Semantic Web application can link together pieces of data from a variety of sources. A user researching a vacation, for example, could type in a query and receive information about all aspects of his or her trip at one time, instead of conducting separate searches for flight, hotel, and car rental information.


The Hunt Is On

The Web can be fooled by keywords. A search on “Troy,” for example, could bring up results ranging from ancient Troy to a university in Alabama to the Troy on the Hudson River. The founders of the search engine Google made a clever end-run around the problem of a keyword in the 1990s, when they organized results using a combination of keywords and human activity on the Web. The top Google results show the pages to which the most people have linked, thus increasing the chances that they are useful. This is not a perfect system, however. Many people may be linking to something other than the page they most want to find.

To see how the Semantic Web would set up alternate methods of finding information, imagine needing to find financial data for a global business, like the airline industry. That information exists on many different Web pages—financial disclosure forms of companies, reports from research analysts, industry newsletters, mainstream news reports, and so on. If you know roughly where to hunt, a sophisticated search engine like Google makes it easier to find those Web pages, but one at a time. The Semantic Web, on the other hand, is designed to extract information from a wide variety of Web pages, by analyzing the nature of the information in those pages.

So if the goal is to find data on the growth in air freight cargo from the United States to China over the last five or 10 years organized by industry or airline, a well-crafted Semantic Web query could link together similar pieces of data from disparate sources. A rich collection of data could quickly be assembled in one place, rather than extracted from a series of painstaking searches. And you probably wouldn’t have to worry about a computer confusing “United,” as in the airline, with “United,” as in “the United States,” among other problems we now face.

“Google is phenomenal, don’t get me wrong,” says Hendler. “And none of this would put Google out of business. What we’re doing with the Semantic Web is building a new link space on top of the old one, so we can link the concepts, not just the documents. So instead of guessing, the minute someone puts that information there, everybody can take advantage of it.” From this perspective, the Web would become, in Berners-Lee’s often-cited phrase, a “Web of meaning,” akin to one giant global database that can be searched, rather than its current form, which is a bit more like one giant pile of file folders to flip through.

* “Untangling the Web”  Page 1 | 2 | 3 | 4    Previous | Next   *
Feature Articles At Rensselaer President’s View Reader Mail Staying Connected Alumni News One Last Thing Back Issues
Rensselaer Polytechnic Institute | About RPI | Academics | Research | Student Life | Admissions | News & Events
Rensselaer (ISSN 0898-1442) is published in Spring, Summer, Fall, and Winter by the Office of Strategic Communications and External Relations, Rensselaer Polytechnic Institute, Troy, NY 12180-3590. Opinions expressed in these pages do not necessarily reflect the views of the editors or the policies of the Institute. ©2008 Rensselaer Polytechnic Institute.