Mike Davidson of Newsvine.com opens by saying this is a time to talk about how to improve the editorial process. How to decide which stories are important and interesting without human intervention? E.g., Techmeme.com looks at what A-Listers and B-Listers are linking to, while Digg lets everyone vote.Newsvine measures how long you spend looking at a story.

Jay Adelson of Digg.com says he’d like to see the mainstream media reflect more of what people actually are interested in. Steven Levy of Newsweek worries that this would result in even more coverage of runaway brides, etc. Digg says that people tend not to digg porn, etc., because it’s associated with their profile. Dan Gillmor wonders how you add reputation to popularity.

Someone asks about journalism on demand. Dan says that some projects are going on now, including Jay Rosen’s NewAssignment.net. Fundable.org also does something like this.

Me: I want to get recommendations based on what my friends are reading and, as Dan points out, what friends of my friends are reading. Jay points to some people’s desire to be anonymous. Dan touts pseudonymity. Karl Fogel says that that permits covert corporate and government sponsorship. Dan asks how we out bad actors. Suggestions: eBay-like recommendation system. Newsvine has a probationary period. Slashdot karma.

Newsvine tried eBay-like ratings — report bad articles — but found that the best writers were about 80% because some people didn’t like what they said. The bland writers had 100%.

Me: The simple way to start is to let me build a list of people I actually know and whose judgment I want to influence what’s recommended to me. Then I don’t have to worry that the person is in fact the CIA or Wal-Mart.

Gabe Riviera of Techmeme is using the implicit social network based on who refers to whom.

Does Diig track how many people diig a page before they’e clicked on the article.Jay doesn’t quite answer.

If we only listen to people we trust, how do we get challenged?

Dan recommends newstrust.net, an effort to measure MSM.

Adrian Holovaty from the WashingtonPost.com is interested in optimizing information collection. How do we get journalists to collect information in ways that machines can reuse it. Newspapers are a collection of information desperate for a framework, while Wikipedia is a framework desperate for information, he says.

Graeme Merrall augments reporters’ stories with metadata.

Dan says there’s a difference between stories and data. Steven Levy says that without training journalists in how to write a story, the data won’t ever become a story.

Already, he says, journalism is becoming a matter of filling in forms and then letting computers build the story. E.g., at one small paper, there’s a visiting band form that the journalists fill in.

Dan points out that Adrian did an app that plotted police/crime info. [I missed the url.]

John Gruber points out that columnists are not so easily replaceable.

Dan rises to defend reporters. Reporting is hard than we’re making out.

Mike Davidson wonders if 5-10 years we’ll be able to say that we want to read a story about the new Apple, written in the style of John Gruber of Steven Levy, etc. He’s skeptical.

Lily Chen says that it depends a lot on what people care about. She cares about what happens on her street but no one is writing about. An automated system might be able to be of value there.

Karl Fogel says that people in the US feel isolated from worldwide news sources in part because there’s no translation. In the open source world, documentation has been translated within days, he says.

Jay wonders if info will continue to go behind the pay wall after a few days. General opinion in the room (actually, in the tent): Nope.

Rabble says that more journalists work as PR people than journalists. Dan says that we need more transparency. Mike of Newsvine says companies have offered to pay them to put their legitimate sources on their site. BestBuy has paid someone to write an article about, say, hot products, that contains a single quote from BestBuy. Newspapers run the article knowing that it’s in effect a paid placement. It’s labeled “ARA” but that’s the only sign.

Adrian says that the categorization onus should be on the reporter. All the info in it ought to be categorized so, if it’s a report on a mayor’s speech, we can see all the speeches by the mayor, all speeches about the same topic, etc.

Graeme points to Visionbytes.com for media search. AOL says that their Drambuie project does something similar. [Tags: journalism digg dan_gilmore newsvine techmeme washingtonpost citizen_journalism everything_is_miscellaneous ]

Categories: Uncategorized dw