Google's Bias for Bigness

By Kelly Hearn, AlterNet. Posted July 14, 2005.

So you need a news fix....

You grab some caffeine, jump online and head to Google News to search for the latest news about, say, New York City's planned Freedom Tower.

It crawls across some 4,600 publications, finds news stories and ranks them from top to bottom.

In this case, the morning of June 29, the page flickers and, bam, Google has found and ranked the stories you need: as its top Freedom Tower story it presents a piece by The Minnesota Star Tribune, followed in downward order by The Los Angeles Times, London Free Press, Xinhua news agency (the official service of the Chinese government).

London Free Press? Xinhua? Does this seem somehow off? Google execs seem to think so, but reports of their proposed fixes raise as many questions as they answer. The company is reportedly concerned about biased or incorrect news stories getting ranked above more quality news. As a fix, they have filed a patent for a new technology that ranks a story by the quality of its source. But for a company that commands a superpower level of control over global Web-surfing behavior, even small changes have a huge effect.

Google doesn't comment on its present or future ranking criteria, but its news stories seem to be ranked based on relevance to search words and timeliness, not the reputation of the news source. But for alternative online news services, from CNET to AlterNet, B2B trade publications, and thousands of small radio and television stations that rely on Google News-related traffic, the company's patent application offers reason to worry about just how shallow a pool of sources the site will draw from. It says the new ranking system will consider:

...[The] average length of an article produced by the news source, an amount of important coverage that the news source produces in a second time period, a breaking news score, an amount of network traffic to the news source, a human opinion of the news source, circulation statistics of the news source, a size of a staff associated with the news source, a number of bureaus associated with the news source, a number of original named entities in a group of articles associated with the news source, a breadth of coverage by the news source, a number of different countries from which network traffic to the news source originates, and a writing style used by the news source.

Would this value-laden code pass judgment on the passive voice? Will it consider contract writers in staff size? Will the "breadth of coverage" criteria take into account that niche-market publications often drill into topics -- be it military procurement or schools funding regulations -- more thoroughly than further-reaching news sources like CNN? Or the fact that these publications often break stories that are then picked up by CNN and others?

Brian Dominick, editor of The NewStandard, an online newspaper that lands a fifth of its daily traffic from Google, writes in his blog that for several reasons the proposed Google News changes would "bury" his publication. He recoils at the idea of assigning a "breaking news" score.

"[That's] like rewarding your boyfriend for climaxing first," Dominick writes. "Breaking news is inherently subject to the most errors and the worst journalism. So it might be good to know that an outlet typically has something early on, but that is not a reliable method for evaluating the quality of its reporting."

Kelly Hearn is a former UPI staff writer who lives in Washington DC and Latin America. His work has appeared in the Christian Science Monitor, American Prospect and other publications.

Liked this story? Get top stories in your inbox each week from AlterNet! Sign up now »