Archive for the 'Knight News Challenge' Category

Getting in metadata game: Oh, money, that’s why!

Who’s mentioned in your article? What organizations does it talk about? Or what zip codes?

Answering these simple questions—in ways notoriously inflexible computers understand—can be like putting handles on your articles. It means aggregators and filterers like EveryBlock can grab on and give readers one more way to find what you have to say.

That’s what the New York Times is doing—in two stages, it appears. First its librarians encode the elected officials mentioned in its articles; mentioning them in the regular text of the article doesn’t cut it. Then its newly built web service, called Represent, figures out the geographic locations those officials represent. Meanwhile, Represent is also taking a computerized look at Congressional votes. When a politician votes, Represent says something like, “Oh, a person just voted in geographic area Y, and that person’s name is X.”

EveryBlock isn’t built for understanding much about people or names, but it is built for understanding locations and geographic areas. So Represent’s job is to translate from X to Y—from names to places.

Which brings us at long last to the metadata game. The historical problem is the way you have to answer these questions has been interminably dull and technical. So the historical result has been one big shoulder shrug: “Why bother?”

Well, people like Adrian Holovaty are starting to envision on answer “We have a number of ideas for sustaining our project,” he writes, “like building a local advertising engine.” That kind of engine might share ad revenue with the newspapers whose articles it incorporates. In order to claim a share, each newspaper must diligently prepare its articles for EveryBlock: there much be location handles that EveryBlock can grab. It’s highly unclear how much money EveryBlock’s hyperlocal ad targeting could generate, but if it’s enough, it will provide the kind of incentive publishers need to make boring metadata worth their while. EveryBlock might just unlock the ‘R’ in ROI. That could very well be a great reason to bother.

Epilogue    It’s notable that grant monies have helped solve this chicken-and-egg problem. I may have personal issues with the Knight News Challenge—I didn’t win and didn’t receive feedback promised on multiple occasions—but EveryBlock is quite justifiably the darling of the news innovation set.

Gatherers and Packagers: When Product and Brand Cleave 4 Realz

Jeff Jarvis writes about the coming economics of news:

When the packager takes up and presents the gatherer’s content in whole and monetizes it—mostly with advertising—they share the revenue. When the gatherer just links, the gatherer monetizes the traffic, likely as part of an ad network as well.

I think this is right. In the first case, the content is on the “packager’s” page or in its feed; in the second, the content is on the “gatherer’s” page or in its feed. In both cases, advertising monetized the content (let’s say) and readers or viewers found it by way of the packager’s brand (a coarse but inevitable word).

To me, however, the location of the user’s experience seems unimportant—in fact, the whole point of disaggregating journalism into two functions, imho, is to free up the content from the chains of fixed locations. Jarvis writes, “The packagers’ job would be to find the best news and information for their audience no matter where it comes from.” I agree, but why not let it go anywhere too—anywhere, that is, where the packager can still monetize it? (See Attributor if that sounds crazy.)

Couple this with the idea that rss-like subscriptions are on the move as the mechanism by which we get our content, replacing search in part. (As has been said before, there’s no spam on twitter. Why not? Followers just unsubscribe.) The result is that the packager still maintains his incentive to burnish his reputation and sell his brand. After all, that’s what sploggers are: packagers without consciences who get traffic via search.

So I agree with Jarvis: “reliably bringing you the best package and feed of news that matters to you from the best sources” is how “news brands survive and succeed.” That’s how “the packagers are now motivated to assure that there are good sources.”

Give me tags, Calais!

Who needs to think about buying tags when Reuters and its newly acquired company are giving them away?

The web service is free for commercial and non-commercial use. We’ve sized the initial release to handle millions of requests per day and will scale it as necessary to support our users.

I mean, Jesus, it’s so exciting and scary (!) all at once:

This metadata gives you the ability to build maps (or graphs or networks) linking documents to people to companies to places to products to events to geographies to … whatever. You can use those maps to improve site navigation, provide contextual syndication, tag and organize your content, create structured folksonomies, filter and de-duplicate news feeds or analyze content to see if it contains what you care about. And, you can share those maps with anyone else in the content ecosystem.

More: “What Calais does sounds simple—what you do with it could be simply amazing.”

If the world were smart, there would be a gold rush to be first to build the killer app. Mine will be for serving the information needs of communities in a democracy—in a word, news. Who’s coming with me?

PS. Good for Reuters. May its bid to locate itself at the pulsing informational center of the semantic web and the future of news prove as ultimately lucrative as it is profoundly socially benevolent.


Josh Young's Facebook profile

What I’m thinking

Error: Twitter did not respond. Please wait a few minutes and refresh this page.

What I'm saving.

RSS What I’m reading.

  • An error has occurred; the feed is probably down. Try again later.