Google Search Needs You Too
Posted by Sam_Gellman on Thursday, Mar 11, 2010
If search is Google’s Big Mac, PageRank is the pattie. But what’s in the special sauce? Steven Levy went inside the Googleplex to get a good look at some of the algorithms and the “signals” (there are 200 of them, including page freshness, your search history, and your location) that leads your search for “hot dog new york” to food, not to underground puppy-boiling dens in the city.
One key: Google uses our searches themselves to tighten up its information machine, and maintain what Levy calls its “relentless march toward search becoming an always-on, ubiquitous presence.”
One unsuccessful search became a legend: Sometime in 2001, Singhal learned of poor results when people typed the name “audrey fino” into the search box. Google kept returning Italian sites praising Audrey Hepburn. (Fino means fine in Italian.) “We realized that this is actually a person’s name,” Singhal says. “But we didn’t have the smarts in the system.”
The Audrey Fino failure led Singhal on a multiyear quest to improve the way the system deals with names — which account for 8 percent of all searches. To crack it, he had to master the black art of “bi-gram breakage” — that is, separating multiple words into discrete units. For instance, “new york” represents two words that go together (a bi-gram). But so would the three words in “new york times,” which clearly indicate a different kind of search. And everything changes when the query is “new york times square.” Humans can make these distinctions instantly, but Google does not have a Brazil-like back room with hundreds of thousands of cubicle jockeys. It relies on algorithms.
And yet, there are hundreds of thousands of people helping the machine get smarter. Maybe billions of people.
At any moment, dozens of these changes are going through a well-oiled testing process. Google employs hundreds of people around the world to sit at their home computer and judge results for various queries, marking whether the tweaks return better or worse results than before. But Google also has a larger army of testers — its billions of users, virtually all of whom are unwittingly participating in its constant quality experiments. Every time engineers want to test a tweak, they run the new algorithm on a tiny percentage of random users, letting the rest of the site’s searchers serve as a massive control group. There are so many changes to measure that Google has discarded the traditional scientific nostrum that only one experiment should be conducted at a time. “On most Google queries, you’re actually in multiple control or experimental groups simultaneously,” says search quality engineer Patrick Riley. Then he corrects himself. “Essentially,” he says, “all the queries are involved in some test.” In other words, just about every time you search on Google, you’re a lab rat.
Filed under:
