Journalism 2.0

A key role for journalism in the digital age is to provide intelligence to data. That emerged as a common theme running through each of the projects developed Friday at GonzoCamp in Seattle. Now that we are swimming in more data than ever, journalists need to develop technology to help them provide intelligence for that data.

– Who owns the company that published a news story?
– What other stories have been reported about this topic?
– What other news occurred in this location?

Those are questions a news organization typically wouldn’t venture to answer. But the independent thinking that powered the first GonzoCamp tackled those questions with a short window for development. I’m impressed by what was accomplished and excited to discuss future GonzoCamp events in other locations to see what else can be built.

We started with some simple questions: What can journalists learn from computer programmers and software developers? And can the programmers and developers learn anything from the journalists?

We also added experienced entrepreneurs and students into the mix and ended up with a surprisingly cohesive group. Based on informal responses from the participants, several important lessons were learned.

“I think it worked great and I would come again,” said Brianne Pruitt, web editor for the Wenatchee World newspaper. “If you work for a news organization, sometime it takes a long time to get anything done. So it was cool to come together today and make so much progress in such a short time.”

With undergrads lounging in the sun near a fountain outside, about 25 people came to GonzoCamp on the campus of the University of Washington. Some traveled more than 100 miles. Some brought rich experience, like Rick Sass, a founding member of USA Today, and Mike Koss, CEO of Faves.com and founder of Startpad.org. Everyone brought ideas and enthusiasm.

The goal was to gather a diverse cross-section of professionals and students together for a day of action instead of words. So there were no presentations, no panel discussions. Just ideas for projects and attempts at progress.

The event, co-sponsored by the UW’s Masters of Communication in Digital Media program and my company, Serra Media, began with about 15 ideas pitched as projects that could be built in a day. None, of course, was going to lift the news industry out of its current financial crisis, but maybe we could discover the seed of an idea that would grow into something much larger. Time will tell.

“Our Master of Communication in Digital Media program is intensely focused on storytelling, social media and new business models of communication,” said Hanson Hosein, a professor at the UW’s MCDM. “That’s why it makes complete sense to host and support GonzoCamp. These journalists, programmers and business development professionals are not only part of our community, they’re at the heart of our mission.”

Idea pitches were followed by votes. Then teams were formed around three ideas. Each team had about 4-5 hours (minus a lunch break for pizza) to build a working prototype that would illustrate the concept. All three teams were able to show something tangible when these ad hoc presentations began around 3:30 p.m. Here’s a rundown:

Seattle News Map: Gregory Heller explained the concept as “sort of an Alltop for local news.” The team, which included Greg Dunlap, Laura Gentry, Jenine Stewart, JP Montagnet, Mike Lewis and Elaine Helm, built a functioning site that aggregates RSS feeds from local news sites and blogs and organizes them by neighborhood. “It’s democratic because anyone can post their own feed, too,” Heller said.

Naturally, the group had more ideas for features than it had time to build, including:
– views for highly rated feeds, most recently updated, most recently added, most visited
– combined feeds by location and topic
– subscriptions and notifications with topic tags
– incorporating Yahoo term extractor that will add terms to the feeds
– other feed sites like Everyblock can be added as a resource
– trending topics view based on tags

“I’m a little disappointed because we lost some time with a minor server meltdown,” Dunlap said, noting that a map interface would have added the “wow” factor the group was hoping for. “But even with our limited time we already have 1,000 articles on the site.”

Still, the concept has some cool implications for hyperlocal news aggregation, an area that continues to garner more attention. One of the best ways for local news sites to extend the capabilities of small staffs is to leverage the power of feeds, aggregation and social selection, core characteristics of the GonzoCamp Seattle News Map prototype.

Whose-news.com: “How diverse is your news consumption?” is a question from the press release (hand-written on spiral notebook paper, of course). Billed as a “nutrition label for your journalism,” the Whose-News technology allows the reader to peel back the layers of a specific news source and reveal the ownership structure behind the publisher.

The Whose-News team, which included Trevor Smith, Rick Sass, Hasan Edain, John Heasly, Seth Long, Monica Guzman and Scott Falconer, built a functional prototype of the technology that would power the concept. The team didn’t have time to build an extensive database of media ownership, but it did build the structure for such a database. And a prototype of the iPhone app.

For example, if you type Wired.com into the search field at Whose-News.com, it returns a page that tells you the owner (Advance Publications) and the outlet (Conde Nast).

“The goal was a bookmarklet in your browser you could use to see who owns news,” Smith explained as he presented the prototype. “The idea is that you’d have a display like a food nutrition label to show the relationship between the owner and the outlet.

“The next step is to get the data and build out database.”

The team had plans for developing an API for “less neutral” sites to suck in the data and do with it whatever they might want. But the idea behind Whose-News is not to judge, but to simply reveal. It would be an easy way for news consumers to see which companies are behind the news they read. It would also come in handy for media criticism blogs and web sites. Much of the data for bigger companies is available from Wikipedia and the Whose-News team mentioned the idea of acquiring the Wikipedia database and parsing the data to build out the Whose-News database.

This is the kind of project that can really only come from a gathering like GonzoCamp. A single news company would have little motivation to build such a tool. But a group of innovative individuals working creatively together can see the benefits for a general news audience and thereby define a simple, but potentially powerful, project.

PagesLike.com: During the morning brainstorming/pitch session, several ideas emerged from a common theme so we tried to put them together with a given team. The GonzoCamp team tackled three ideas all related to tagging or classification of news and came out with, well …

“This is none of those 3 ideas,” Roy Leban said in presenting a prototype for what the group did come up with: PagesLike.com.

The idea is to provide a news site with technology that says “if you like this story you are reading, then you might like these other stories.” Those “other” stories could come from any Web site, making PagesLike.com a valuable form of automated link journalism.

The concept is to have one line of Javascript code in a page template do it all. You can see the working prototype in the right column at blog.pageslike.com. The system would gain intelligence over time and record keywords that appear repeatedly in news stories.

The team, which included Chuck Taylor, Mike Koss, Dylan Wilbanks, Jonathon Fitzpatrick, Aaron Ritchey, Pruitt and Leban, spent most of the morning (and some of the afternoon) fine-tuning the idea. so the first line of code wasn’t written until about 1:30, yet the widget was operational by 4 p.m. Plus the team created a blog (and Twitter account) for good measure, where Pruitt posted the following explanation of the service:

Create context and relevance at the article level by providing your readers with links to other Web sites with similar content. No work on your end, other than embedding the widget, and a lot of added value to your readers. You provide the context, the other perspectives to the story, by linking to outside sites. And we do it all for you. Our program will search your articles for keywords and then provide a list of links to articles on other Web sites that have the same keywords. The more sites that embed our widget, the more diverse the links you provide.

This area of link journalism and aggregation is gaining momentum as activity surrounding the so-called semantic web picks up steam. Some sites have developed or deployed technology that aims to serve a similar objective as PagesLike.com. For example, the Washington Post is using Evri. Plus, there’s the OpenCalais project from Thomson Reuters.

* * *

Each of the teams is responsible for deciding if – and how – to continue the projects. It won’t be easy, of course, to juggle day jobs with more side projects. But hopefully the seeds planted Friday will get a chance to grow and blossom and lead to other interesting ideas and concepts.

UPDATE: See what Greg Dunlap wrote about his impression of GonzoCamp on his blog: Heyrocker.