I just heard an excellent presentation by Krishna Bharat of Google News. He explained how Google News works. It basically crawls news sites, finds "story clusters", ranks the sources, figures out how prominently each source is running the story, figures out whether its a big story or a little story, figures out geographic references, and builds the pages for the various geographic and language editions. He was talking to an audience of editors so there were many questions about how the "editing" process worked and many people couldn't seem to believe it was algorithmic. Some people seemed afraid that Google News would replace them. The point that he made and was clear from the process that he explained is that it uses the decisions that the editors of the various media make about what story to run and where in deciding how important a story was. It was basically aggregating the decisions of the editors, not replacing them. Without the editors and the "front page process" Google News couldn't decide what story to lead with. At least in its current form.

The derivative conclusion you can come to is that Google News is just amplifying or reinforcing systemic biases in MSM editorial and NOT helping to address these issues. I think this make Google News very news media friendly and also provides an opportunity for bloggers and projects like Global Voices to still have a very important role. I guess that if Google New started incorporating more of the alternative press, they could shift the bias.

During the discussion, Dan Gillmor pressed Krishna for more transparency on the algorithm and the list of sources and I seconded the motion.

Some good notes of the sessions on the editors blog.

4 Comments

Google's founders seem to think that when something is benign it is ok. That is their definition of not "evil." In their little world censorship in China is ok. It is also ok to regurgitate other organizations "news" without editorial comment or concern. "We don't write the shit, we just barf it back at our customers who want it." I do not use Google news because it is news without a soul. Even Fox News has a human deciding what is news and what isn't. I am a technologist but I do not want an alogrithm choosing the top story for me. It is a form of the tyranny of the masses, the basic concept that the majority can be wrong, sometimes with great consequences. Organizations that facilitate the tyranny of the masses are "evil" in my mind.

This is the one issue about Google that actually worries me. Recently I heard Eric Schmidt (Google CEO) speak here in LA and one question that was consistently raised by the prominent bloggers in the audience was that all of google's search statistics were kept secret. That information is extremely valuable to a lot of people so what Google does with it is indicative of the direction we are all headed in. Eric essentially answered that the data would remain proprietary but that they would use it to consistently improve search results.

Now, just a couple of months later, Current.tv (Al Gore's new youth TV network) has announced that they will be getting the latest in Google's stats to use for their on-air programming. In other words, Al Gore is paying Google to get access to this proprietary data which they will then use to craft what kind of shows they think are topical. This is very scary people. Why doesn't google make this data available as an API? This info is too valuable to make available to the highest bidder.

I came across another similar site to Google news. Check it out at http://news.kify.com

Not as sophisticated as google news, but worth watching it.

Leave a comment

1 TrackBacks

Listed below are links to blogs that reference this entry: Inside Google News.

TrackBack URL for this entry: http://joi.ito.com/MT-4.35-en/mt-tb.cgi/3525

Joi Ito weiss mehr und auch Dan Gilmore war dabei, ebenso wie Andrew Nachison vom editorsweblog.org. In einer Session auf dem World Editor's Forum hat der Erfinder von Google News Krishna Bharat ein wenig mehr dar Read More

About this Archive

This page is an archive of recent entries in the Business and the Economy category.

Books is the previous category.

Computer and Network Risks is the next category.

Find recent content on the main index.

Monthly Archives