Teradata Magazine Online

Connections

Next

Promise of a
better Internet

The semantic Web could enable even greater access to information.

by Elizabeth Millard

During the past 15 years, the nature of the World Wide Web has shifted dramatically—from a static source of information similar to newspapers, bulletin boards and printed letters (Web 1.0) to a vibrant community exchange place (Web 2.0). Capabilities such as social networking, wikis and instant messaging have transformed the way people and enterprises use the Web, making it a vital part of their lives.

But despite the relative newness of Web 2.0, another revolution may be brewing. The World Wide Web Consortium (W3C)—a major driver of Web innovation—believes the semantic Web, also called Web 3.0, is on the horizon and that this shift could make online information even easier to find and use.

The semantic Web

With existing data structures and networks, informa­tion needs a high degree of manual manipulation, despite some automated storage and retrieval processes. IT managers, data warehouse managers and others must direct data according to a company’s data manage­ment processes and policies. The semantic Web, however, could significantly change how data is handled.

Currently, computers can find information stored in documents and display that data, but they can’t interpret how relevant that infor­mation might be, and they can’t act on it intelligently. In the semantic Web model, data itself becomes part of the Web so that it can be processed independently of platform or application.

Compare this with using a search engine: Certain keyword combi­nations bring up results, but the computer doesn’t examine a user’s most recent documents, online searches and data stores to determine the context of that search. If it did, it could produce a web of data that would operate under a common language and return a single, highly customized search result.

For data collection efforts, this would eliminate many of the soft­ware interoperability issues that plague information retrieval. Instead of tucking data into proprietary applications—and running the risk of not being able to access the data because of software issues—the semantic Web would put all data into the same format, with the same rules, so it would operate like a single, sizable online application with raw data including text, video and images.

For instance, an Internet user might have bank statements, photos and appointments online in different applications. These types of data can’t easily come together now, but with the semantic Web, the user might be able to see bank charges on a certain date or whether certain photos were taken in a specific month, simply by using a single search query or by seeing the information and images linked together in a different graphical interface that would be developed later when the semantic Web has been adopted.

Development and vision

Although the concept is cutting-edge, the semantic Web has been in the works for nearly a decade. First conceived by Web pioneer and WC3 Director Tim Berners-Lee, the semantic Web was imagined as “a web of data that can be processed directly and indirectly by machines.” Berners-Lee believed the Web could evolve into a universal medium for information, data and knowledge exchange, rather than just being a repository for informa­tion or a mode of communication.

To move toward this vision, collaborative working groups have harnessed enabling technologies and tinkered with a set of design principles and specifications. One of the most important is the resource description framework (RDF), originally conceived as a metadata model but which came to be used as a general information modeling specification. Once data is in RDF form, it becomes easier to process, because RDF is a generic format and the information that is contained in it maps directly and unambiguously to a decentralized model. Developers and programmers dealing with RDF applications can quickly see which bits of data fit in with the semantics of an application once they learn the complexities of the framework.

Not surprisingly, the transition to the semantic Web faces numerous hurdles. Beyond technical issues, such as dealing with the complexity of a graph-based language like RDF and tackling the annotation of infor­mation across the Web, critics have questioned whether such a system could adequately protect privacy or whether it could even be used for censorship. If information is placed into an all-encompassing data web, it would be difficult for users to maintain anonymity or to stop others from viewing online information they’ve created.

Impact on data warehousing

Because the semantic Web would have an effect on how data is handled, data warehousing would be affected by extension. In essence, removing a certain level of human involvement in data processing could allow machines to communicate and collabo­rate more effectively with one another, so that data could be organized more efficiently and discovered with less effort.

That type of system might decrease the need for time-consuming information asset tracking and data management and could even affect data volume, because intelligent agents could work to scrub out superfluous and duplicate data from a warehouse. Also, the semantic Web could spur better customer interactions through faster, more reliable searches.

For example, the semantic Web could be used for a product sold over the Internet, with the system possessing two main applications: one for people who wanted to buy the product and another for people selling the product. Each buyer’s information would be stored in an RDF file, and product searches would involve a Web spider continu­ously searching for relevant RDF files from sellers. Unlike current buyer and seller systems, the system wouldn’t rely on as much human interaction for searches but would, instead, match RDFs.

Although there are no data warehousing projects that are widely publicized yet—companies are waiting for the working groups to hammer out the technologies and standards associated with the semantic Web—there’s certainly a great deal of promise behind the shift. Semantic Web researchers are working to create more RDF documents so early adopters will have more opportunity to play with scalability and usability. And, as Berners-Lee has noted, the Internet infrastructure to power the semantic Web already exists.

Even now, there is consumer application development based on the semantic Web idea, including contextual search, browsing tools and personal information management systems. With all of the develop­ment in place and a strong grassroots effort from working groups, the shift to the semantic Web shouldn’t take another 15 years.

Elizabeth Millard lives in Minneapolis and writes about business and technology.