AJR  Features :     * WEB ONLY    
From AJR,   August/September 2008

Welcome to Web 3.0   

Online Exclusive Getting to Know the Semantic Web

By James Sanborn

Just as news publishers come to terms with Web 2.0, along comes Web 3.0 to shake things up.

With Web 2.0, the focus is self-publishing and connectivity. Think YouTube and blogging, Facebook and podcasting.

Web 3.0, also known as the Semantic Web, uses smart programs to tag and link to information across mediums, providing context and depth to stories without much human intervention.

The fully realized Semantic Web is a vision on the horizon. But it's one many online leaders are taking seriously.

It could bring more precise search results to users, drive traffic to smaller local publications and change the way all online publishers reach their audiences.

One way to keep people on a news site is to serve as an information hub, says Tristan Harris, CEO of the software firm Apture

In the Semantic Web, "news lives in an ecosystem," Devin Wenig, CEO of Thomson Reuters' Markets Division, said during a panel at this month's Online News Association conference in Washington, D.C. It meets the demands of users who "seek context and a 360-degree view of issues," he said.

ClearForest, a Thomson Reuters subsidiary, is designing software that helps lay the ground floor for the Semantic Web. It connects related information across platforms without the hand of an editor--a defining characteristic of this Web, which is the brainchild of Tim Berners-Lee, creator of the World Wide Web.

In May, ClearForest released Calais 2.0, a free program that scans content and suggests meta tags that computers can read to automate the process of relating and linking information by topic. Publishers can submit a story to Calais, and in a split second the service spits back a list of terms categorized by people, companies, places and events. Used as meta tags coded into the top of a story, those terms can have profound effects on readers and publishers alike. For publishers, it helps put content high in search engine results. For readers, it makes results more tailored and relevant.

While it's clear the Semantic Web is an important development to watch, Doug Feaver, a washingtonpost.com blogger and the site's former executive editor, cautioned that people shouldn't leap into the new technology before becoming fully informed.

"I worry if we understand exactly what the implications may be," he says. "Even when you go to Google you don't always really know what you are getting back. What is the real source? What kind of connections were applied?"

"Computers sometimes make interesting connections that aren't quite right," he says.

The Semantic Web could help expand the audiences of smaller and local publications, but not without growing pains.

Better search functionality helps local publications increase traffic, says Dale Steinke, interactive news and operations manager for KING5.com, a Seattle Web site.

Now that his site works to extensively label content, he has seen an uptick in traffic from the Seattle area and around the world, a trend that will likely grow as the Semantic Web proliferates.

KING5.com might not seem like the logical place to go for Texas hurricane coverage. But if content is labeled well, readers might type in Hurricane Ike and get KING5.com's coverage near the top of their search engine results.

The downside, Steinke says, is that although the Semantic Web will open the door to a wider audience, it could also fracture it. People are going less and less to news home pages to take in the day's news. They are increasingly using search engines to find particular stories inside a site, another trend the Semantic Web will encourage.

"It's less likely that people will specifically come to KING5.com," Steinke says. "They will use a tool to search for a story and that tool will direct them to a Web page, hopefully ours."

That means sites have to rethink their design, Steinke says. "Our job is to now create an atmosphere around those pages that keeps people on our sites longer," he says.

One way to keep people on a site is to serve as an information hub, says Tristan Harris, CEO of the software firm Apture. ProPublica, washingtonpost.com, the BBC and the Wall Street Journal's health blog use Apture to connect users to relevant information from across the Web while keeping them on their sites.

By harnessing the Semantic Web's ability to automatically associate related content across mediums, an editor can use Apture to highlight any term on a page--Myanmar, for instance. Apture will automatically use "Myanmar" to find related videos, Wikipedia entries, audio recordings, PDF documents or even in-house content. With just a few clicks of the mouse, an editor can choose what content to associate with that link.

When a user clicks on links, Apture opens them in smaller windows that float above a page. A user reading about strife in Myanmar can quickly learn about the country's history, politics, what it looks like, what language its people speak and what music they listen to, all without ever leaving a Web site. It satisfies reader curiosity while keeping them on a page. And longer user times on a site can provide financial benefits when dealing with advertisers.

Harris believes that by serving as an informational hub, a site will draw loyal and repeat users. The days of simply presenting your own content are over, he believes.

"Pretending that you are the only one [who] can tell a story is an old idea," Harris says. "If you continue with it, people are going to go away."

Sanborn is a writer for the Philip Merrill College of Journalism's student Web site Maryland Newsline.