TECHNOLOGY/WIRELESS
IN DEPTH: NEW XML STANDARD TAGS NEWS YOU CAN REALLY USE
By Liz Garone

(8/3/00) Anyone who has ever tried single-handedly to turn a straight, print-based news story into an interactive Web experience--or to repackage a feature for several different markets or distribution media--has reason to take heart. The recent announcement by the International Press Telecommunications Council (IPTC) that it is adopting a standard for formatting electronic content--a subset of extensible markup language (XML) (see "In Depth: When Is a Language Not a Language?" February 2, 2000) called NewsML--offers a glimpse of better things to come.

"This is the Holy Grail--to have a widely-supported XML-based standard for shipping data around that you don't have to go through and artificially clean up, [hoping] that you got all of the jump lines [which indicate where a story is continued] out of the story," says an ecstatic Steve Yelvington, a private online media consultant.

Yelvington knows first-hand the effort that goes into turning a print product into an Internet package--he is the former executive editor of Cox Interactive Media in Atlanta.

"The core of the problem is that for a long time news has been shipped around in some very naive packaging. The current standard for transmitting wire stories really dates back to the years of lead typesetting and letter-press printing and isn't very sophisticated," he says. In an era where features are produced from a core set of data and news is delivered via cell phone and PDA, he says, "we really need information to be structured in a way that is both machine-readable and reflects a higher level of organization than what we've had in the past."

Marking up the news

NewsML is poised to transform the delivery of news, according to Yelvington. As an open-source standard, NewsML can be used for all aspects of multimedia news creation, storage, and delivery, on any platform and on devices from PCs to cell phones and PDAs.

At NewsML's core is the NewsItem, which can contain different media (text, photos, graphics, audio, video). The NewsItem also includes meta-information that explains the relationship between, and role of, each of the item's components. This allows for text in different languages, as well as video clips and images in different formats to be used and reused in various media without recoding.

"It's an XML-based way of describing how things fit together, and it builds on top of other standards but really gets closer to what we needed to have," Yelvington says. "What we're really trying to get is information that is marked up in ways that make it more useful."

NewsML uses standard Internet naming conventions for identifying the news objects in a NewsItem. Content doesn't have to be embedded in a NewsItem. Instead, pointers can be used so that users get data only when they need it, making NewsML more bandwidth-efficient than its previous standards.

Repurposing and narrowcasting NewsML will open up new avenues for distributing existing content, says Sebastian Holst, vice president of marketing for Artesia Technologies, which manages digital content for General Motors and The Washington Post.

"Knowledge and information and digital assets will have application and value across what have been traditionally distinct market segments," Holst says. "Just because something shows up originally as a news item doesn't necessarily mean that I wouldn't want to be able to search and re-use and perhaps license some of the imagery or content in a documentary or in an online encyclopedia."

One area that Yelvington is especially excited about is the geocoding and micro-zoning news. He envisions pulling out a GPS-outfitted Palm-style device, searching for, say, entertainment news, and then receiving items relevant to the geographic area in which he happens to be. Issues of time will be circumvented through use of a universal time stamp.

NewsML also opens up the new stores of digital content to small, tightly focused news and information services. "The opportunity is now there for anyone who is downstream from the information wholesalers, regardless of whether you're AOL/Netscape or you're a smaller player with a smaller focus," Yelvington says. "If you have access to information that's highly structured, then you can do interesting things with it."

Holst agrees. "[Smaller news purveyors] can certainly contribute and bid on licensed information much more readily," he says. "If they want to provide related stories or supplemental information or incremental research, they now have a mechanism to tie their assets into the broader collection of content, which previously wouldn't have been possible or would have been very, very expensive."

The contrarian view

Not everyone, however, is greeting NewsML with enthusiasm. David Galbraith, vice president of strategy and founder of Moreover.com, which bills itself as the world's most comprehensive online news aggregation service, is taking a wait-and-see attitude. "People are trying to create standards for these things, but I haven't seen one yet that's succeeded," Galbraith says.

He continues, "What tended to happen was people were over-ambitious to start with, and some of the very, very simple XML vocabularies have been more widely adopted than the really ambitious ones. As I understand, NewsML is a compromise between simplicity and comprehensiveness." If NewsML doesn't prove its worth, that's OK, too, Galbraith says: "It's completely Darwinian. The best standards will win."

But whether it's NewsML or something else, having an across-the-board standard is imperative--and on this, Holst and Galbraith agree. "If all these different industries had totally different ways of looking at the same information, it would be impossible," Holst says. "This [announcement] is good news. They're not just using the same standard--XML--and defining standards within their group. They're laying the groundwork for a much more fluid sharing and re-using of digital assets."

Reuters plans to demonstrate NewsML on its site next month, and the IPTC plans to approve it officially in October. The beta version of NewsML is available on the IPTC Web site. <<

< Back