Berkeley CSUA MOTD:Entry 33729
Berkeley CSUA MOTD
2019/03/20 [General] UID:1000 Activity:popular

2004/9/23 [ERROR, uid:33729, category id '18005#12.8525' has no name! , ] UID:33729 Activity:kinda low
9/23    The unintentional right wing bias of the Google News algorithm
        \_ wtf?  last week I said it was left wing!
Cache (8192 bytes)
Have you observed any discernible bias in news aggregators such as Google News or Yahoo News? Do you think human editors do a better or worse job of selecting news stories than computers do? Posted: 2004-09-23 In newspaper newsrooms, editors often go to great lengths to achieve a semblance of balance in coverage of the two major candidates for president. Others make sure that photo size and placement don't favor one over the other. But what are the rules for online search engines, where millions of users are turning for their daily news fix? Does evenhanded coverage apply in the bottomless news hole of cyberspace? Does having an editorial team or an automated program get you a better sweep of important news about the political candidates? To their credit, Google News and Yahoo News agreed to pull back the curtain and explain how they acquire and display political news. Google News now attracts about 6 million users a month, double the audience of a year ago. In August it drew 58 million visitors, making it the 14th most popular site on the Web for current events and global news, according to Nielsen//NetRatings. Google News scours not the entire Web but 7,000 information sources (4,500 of them English-speaking) and then groups and prioritizes the news into clusters of articles. An internal "sourcing team" decides which information providers to comb, but for competitive reasons Google would not disclose which sources it uses. Despite those predictable flaws, it's been puzzling to read Google News' takes on John Kerry and George W Bush over the past month. com: John Kerry Just Can't Stand The Heat In addition to mainstream news outlets from both sides of the political fence (say, NPR and The Washington Post on the left and The Washington Times and New York Post on the right), there were 34 anti-Kerry screeds from the second-tier websites. Far from an isolated example, the pattern has repeated itself throughout the past month. Small conservative Web sites such as Useless-Knowledge, Men's News Daily, Michnews and ChronWatch turn up in disproportionate numbers when clicking on news about John Kerry. Useless-Knowledge, for instance, made up 12 of the first 100 results for John Kerry on Friday, and 11 of the first 100 results Saturday. George W Bush typically results in a fairly neutral, evenly balanced set of results from both sides of the political spectrum, with many of the same small conservative sites showing up to sing the president's praises. Krishna Bharat, chief scientist for Google News, said he was puzzled by reports that the service has been skewing politically in one direction. "Google News is a bit like a conversation that we're hosting," he said by phone from India. "We're inviting thousands of news sources to take part, even those who are very small. The two big things we're seeking are inclusion -- we want everyone at the table -- and diversity of opinions in the press." Bharat said Google News uses a mix of techniques to ensure that users are presented a diverse range of perspectives. The ranking and prominence of stories are based on several factors: How many publications are writing about a topic; the size of the story, with substantive pieces ranking higher than short items; and the frequency of the search term within the article. The computer algorithms, he said, "are trying to understand how hot and how big the story is." Every 15 minutes a new edition of Google News is generated and the ranking changes. The formula rearranges the headline blurbs in each story cluster based on the freshness of each article and the importance of the source. "The algorithms do not understand which sources are right-leaning or left-leaning," Bharat said. Google News does not use the same formula as Google's general search engine, which ranks results based on how many people are linking to a site or article. Bharat points out, however, that link popularity plays no result in Google News' rankings. I would like Republicans and Democrats alike to read pro-Kerry and anti-Kerry articles, but it's not our job to change the natural range of opinions that you see in the press. Why does clicking on a "John Kerry" link in Google News turn up so many second-tier conservative sites but so few liberal sites? Bharat said it might be an aberration, and that more people might type in "Kerry," which gives you a more balanced set of results, drawing more articles from major media organizations. But that ignores the fact that Google News itself uses "John Kerry" as the preferred search term when it highlights news about the candidate in its In the News section. A search for "Kerry" on Google News turns up mostly mainstream media sources, while a search for "John-Kerry" -- the search conducted when you follow the In the News link -- turns up a great deal more opinionated pieces culled from second-tier and fringe sites. com and now a fellow at Harvard's Berkman Center for Internet and Society who studies search engines. The chief culprit, he theorized, is that mainstream news publications refer to the senator on second reference as Kerry, while alternative news sites often use the phrase "John Kerry" multiple times, for effect or derision. A second possible factor, Zuckerman said, is that small, alternative news sites have no hesitancy about using "John Kerry" in a headline, while most mainstream news sites eschew first names in headlines. The inadvertent result is that the smaller sites score better results with the search engines. "You have to wonder why some of these wacky sites make the cut," he added. With an occasional exception, Weblogs are generally not found among the Google News results, so Zuckerman had some advice for aspiring political publishers who want to game the search engines: Don't blog -- start an alternative news network. Use terms like George Bush and John Kerry frequently, rather than their last names alone, in both your text and headlines. What Zuckerman calls gaming the system, others call optimizing your site. Thomas Krafft, a Web site developer, said he began working with the conservative news site ChronWatch nearly three years ago when it was averaging 100 visits a day. "I completely rebuilt the site to better organize, categorize and display the content, to ease the process of adding articles to the site, and to especially be more search-engine friendly," he said by e-mail. "Today, ChronWatch averages nearly 10,000 visits per day and is regularly placed near the top of the Google News service," Kraft said. "And it looks like we're on track to see the same results and popularity for the site through MSN's new search service as well." Krafft said he also advises his clients to use keywords and phrases that match users' precise searches and to write in informal, accessible language. ChronWatch editor Jim Sparkman said the site attracts volunteer contributors who believe passionately in their cause. "There are many, many sites like ours on the Internet, and many, many e-mail exchange groups, that are forming a new communication method that is beginning to rival the big media in influence," he said. In August, Yahoo News attracted 22 million unique visitors to CNN's 229 million, according to Nielsen//NetRatings. Jeff Birkeland, product manager for news, said Yahoo is out "to create the broadest, deepest, most comprehensive and useful news experience from start to finish." Toward that end, Yahoo News hosts breaking news, features and analysis from more than 100 news partners, mostly major news organizations. Election 2004 include breaking news from partner news sources, pointers to political blogs and the candidates' sites, and in-depth analysis and commentary. Readers who want to go deeper can plumb Yahoo's news search, which indexes more than 8,000 sources. Like Google News, Yahoo won't disclose how a term like John Kerry or George Bush makes it to the front page of its search results, but Birkeland said the factors include the source, the freshness of the story, and a method of determining relevance. Yahoo achieves balance in political coverage by using a wide variety of news partners and an editorial staff that pulls together "a v...