"Long before it's in the papers"
June 04, 2013

RETURN TO THE WORLD SCIENCE HOME PAGE


Google’s kinship with the mind

Dec. 5, 2007
Special to World Science  

Think­ing and us­ing an In­ter­net search en­gine might seem to be two very dif­fer­ent ac­ti­vi­ties. But a study sug­gests they draw on si­m­i­lar prin­ci­ples.

When you type words in­to the pop­u­lar Goo­gle search en­gine, it re­turns a list of web­pages con­tain­ing those words. The list is­n’t or­dered any old way: it’s or­dered based on how “im­por­tant” Goo­gle deems the pages to be. Goo­gle meas­ures a page’s “im­por­tance” us­ing a for­mu­la based on pop­u­lar­ity. It takes in­to ac­count how many oth­er pages link to that page; how many oth­ers, in turn, link to those; and so on.

(Courtesy Google Inc.)


Now, psy­chol­o­gists have found that our brains re­turn re­sults in much the same way when giv­en a sim­ple task, such as to think of a list of words that start with A.

Thom­as Grif­fiths of the Un­ivers­ity of Cal­i­for­nia, Ber­k­e­ley, and col­leagues ranked the “im­por­tance” of over 5,000 words us­ing the same basic Goo­gle for­mu­la, called Page­Rank. But in­stead of In­ter­net links, the re­search­ers tal­lied men­tal “links” be­tween words as re­flected in an­swers giv­en in word-associa­t­ion games by peo­ple par­ti­ci­pat­ing in pre­vi­ous stud­ies.

The investigators found that a word’s “Page­Rank” was a good pre­dic­tor of how of­ten it would show up when peo­ple were asked to think of words that start with A, with B, and so on.

When it came pre­dict­ing these re­sults, “Page­Rank” beat two oth­er seem­ingly rea­son­a­ble rank­ing sys­tems: tal­lies of how of­ten words show up in or­di­nary writ­ing; and a sim­ple count of di­rect “links” to a word that does­n’t con­sid­er how many words, in turn, link to those link­ing words.

In the PageR­ank for­mu­la, a page gains “im­por­tance” based on how many oth­er pages link to it. But links from pages that are them­selves “im­por­tant,” con­fer more im­por­tance than those that aren’t. Thus, im­por­tance can be thought of as flow­ing through the Web’s link net­work to­ward the most highly “linked-in” sites.

One ex­plana­t­ion for the new find­ings, wrote Grif­fiths and col­leagues, could be that con­nec­tions among brain cells work si­m­i­larly to Web links. Cells that are tar­gets of many con­nec­tions might be­come more ac­tive than oth­ers, in the same way that highly linked-in web­sites are deemed more im­por­tant.

“Our ap­proach in­di­cates how one can ob­tain nov­el mod­els of hu­man mem­o­ry by stu­dying the prop­er­ties of suc­cess­ful in­forma­t­ion-retrieval sys­tems, such as In­ter­net search en­gines,” the group wrote in the stu­dy, pub­lished in this mon­th’s is­sue of the re­search jour­nal Psy­cho­log­i­cal Sci­ence. The study al­so sug­gests brain sci­ence might help de­sign bet­ter search en­gines and data-retrieval sys­tems, they added. “These prob­lems are ac­tively be­ing ex­plored in com­put­er sci­ence,” they wrote, but “one might be equally likely to find good so­lu­tions by stu­dying the mind.”


* * *

Send us a comment on this story, or send it to a friend

Reference: TL Griffiths, M Steyvers, A Firl, 2007. Google and the Mind: Predicting Fluency With PageRank. Psychol. Sci. 18, 1069-76.


 

Sign up for
e-newsletter
   
 
subscribe
 
cancel

On Home Page         

LATEST

  • Meet­ing on­line may lead to hap­pier mar­riages

  • Pov­erty re­duction, environ­mental safe­guards go hand in hand: UN re­port

EXCLUSIVES

  • Was black­mail essen­tial for marr­iage to evolve?

  • Plu­to has even cold­er “twin” of sim­ilar size, studies find

  • Could simple an­ger have taught people to coop­erate?

  • Diff­erent cul­tures’ mu­sic matches their spe­ech styles, study finds

MORE NEWS

  • F­rog said to de­scribe its home through song

  • Even r­ats will lend a help­ing paw: study

  • D­rug may undo aging-assoc­iated brain changes in ani­mals

Thinking and using an Internet search engine might seem to be two quite different activities, but a study has found they may rely on similar principles. When you type words into the popular Google search engine, it returns a list of webpages containing those words. The list isn’t ordered any old way: it’s ordered based on how “important” Google deems the pages to be. Google measures a page’s “importance” using a formula based on popularity. It takes into account how many other pages link to that page; how many pages, in turn, link to those; and so on. Now, psychologists have found that our brains may return results in much the same way when given a simple task, such as to provide a list of words that start with A. Thomas Griffiths of the University of California, Berkeley, and colleagues ranked the “importance” of over 5,000 words, using the same Google formula, called PageRank. But instead of Internet links, the researchers tallied mental “links” between words as reflected in answers given in word-association games by people participating in previous studies. The researchers found that a word’s “PageRank” was a good predictor of how often it would show up when people were asked to think of words that start with A, with B, and so on. When it came predicting these results, “PageRank” beat two other seemingly reasonable ranking systems: tallies of how often words show up in ordinary writing; and a simple count of direct “links” to a word that doesn’t consider how many words, in turn, link to those linking words. In the PageRank formula, a page gains “importance” based on how many other pages link to it. But links from pages that are themselves “important,” confer more importance than those that aren’t. Thus, importance can be thought of as flowing through the Web’s link network. One explanation for the new findings, wrote Griffiths and colleagues, is that connections among brain cells work similarly to Web links. Cells that are targets of many connections might become more active than others, in the same way that highly linked-in websites are deemed more important. “Our approach indicates how one can obtain novel models of human memory by studying the properties of successful information-retrieval systems, such as Internet search engines,” they wrote in the study, published in this month’s issue of the research journal Psychological Science. The study also suggests brain science might help design better search engines and data-retrieval systems, they added. “These problems are actively being explored in computer science,” they wrote, but “one might be equally likely to find good solutions by studying the mind.”