Information Foraging Theory: Adaptive Interaction with Information PDF
Preview Information Foraging Theory: Adaptive Interaction with Information
INFORMATION FORAGING THEORY OXFORD SERIES IN HUMAN-TECHNOLOGY INTERACTION Series Editor: Alex Kirlik Adaptive Perspectives on Human-Technology Interaction: Methods and Models for Cognitive Engineering and Human-Computer Interaction Edited by Alex Kirlik Computers, Phones, and the Internet: Domesticating Information Technology Edited by Robert Kraut, Malcolm Brynin, and Sara Kiesler Information Foraging Theory: Adaptive Interaction with Information Peter Pirolli Information Foraging Theory Adaptive Interaction with Information Peter Pirolli 1 2007 1 OxfordUniversityPress,Inc.,publishesworksthatfurther OxfordUniversity’sobjectiveofexcellence inresearch,scholarship,andeducation. Oxford NewYork Auckland CapeTown DaresSalaam HongKong Karachi KualaLumpur Madrid Melbourne MexicoCity Nairobi NewDelhi Shanghai Taipei Toronto Withofficesin Argentina Austria Brazil Chile CzechRepublic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore SouthKorea Switzerland Thailand Turkey Ukraine Vietnam Copyright(cid:1)2007byPeterPirolli PublishedbyOxfordUniversityPress,Inc. 198MadisonAvenue,NewYork,NewYork10016 www.oup.com OxfordisaregisteredtrademarkofOxfordUniversityPress Allrightsreserved.Nopartofthispublicationmaybereproduced, storedinaretrievalsystem,ortransmitted,inanyformorbyanymeans, electronic,mechanical,photocopying,recording,orotherwise, withoutthepriorpermissionofOxfordUniversityPress. LibraryofCongressCataloging-in-PublicationData Pirolli,Peter. Informationforagingtheory:adaptiveinteractionwithinformation/PeterPirolli. p.cm.—(Oxfordseriesinhuman-technologyinteraction) ISBN978-0-19-517332-1 1. Informationbehavior. 2. Human-computerinteraction. I. Title. II. Series. ZA3075.P572007 025.5'24—dc22 2006021795 9 8 7 6 5 4 3 2 1 PrintedintheUnitedStatesofAmerica onacid-freepaper To my lifelong partner Jacqui, and my parents Remo and Betty. This could not have happened without your love. This page intentionally left blank Preface Tounderstandtheevolutionofthings,onemustun- studies of people doing information-intensive work derstandsomethingabouttheirhistoryaswellasthe started to give me some sense of the range of phe- environmentalforcesthathadshapinginfluencesupon nomena that we would need to address. Our study them.InformationForagingTheoryevolvedthrough participantsclearlywerefacedwithmassivevolumes aseriesoffortuitoushistoricalaccidents,aswellasa ofinformation,oftenunderdeadlineconditions,and number of enduring shaping forces. A critical event making complex search decisions based on assess- was my move to the Palo Alto Research Center ments that were enveloped in a great deal of un- (PARC).SoonafterIcametoPARCatthebeginning certainty. of1992,Ibecameinvolvedintryingtodevelopstudies These information-intensive tasks seemed to be andmodelsaroundasetofprojectsthatwerecollec- different than the human-computer interaction tasks tively called intelligent information access. This in- that were being addressed by cognitive engineering cluded the novel information visualization systems models in the early 1990s, or the science, math, and investigatedintheUserInterfaceResearchArea(see, programming tasks addressed by intelligent tutoring e.g.,Cardetal.,1999)aswellasthenewtechniques systems of that same period. Such cognitive models for browsing and searching being created in the addressed tasks that tended to occur in task environ- QuantitativeContentArea(e.g.,Raoetal.,1995).As ments that (although large and complex) were well partofthiseffort,agroupofus(includingStuCard, definedbya circumscribeddomainofpossible goals, DanRussell,MarkStefik,andJohnvanGigchfrom elements of domain knowledge (e.g., about Lisp pro- California State University—Sacramento) were run- gramming, algebra, word processing), and potential ning some quick-and-dirty studies of people such as actions (e.g.,ina formallanguage,or ina userinter- businessintelligenceanalystsandMBAstudents.Our face). In contrast, the behavior of people seeking viii Preface information appeared to be largely shaped by the of reactions. The first was one of disbelief in the structure or architecture of the content—the in- analogy, for a variety of relatively good reasons (e.g., formation environment—and only minimally shaped humans are not rational, information is not food). bytheuser’sknowledgeofuserinterface.Inaddition, The second was that the ideas were ‘‘audacious’’ (to the structure of the information environment was quote Jock Mackinlay). Fortunately, Stu Card (my fundamentally probabilistic. Consequently, behavior manager and colleague in the User Interface Re- was also dominated by choices made in the face of searchArea)pushedmetopursuethisapproach,and uncertainty and the continual evaluation of the ex- he has been my main sounding board for the devel- pectedcostsandbenefitsof variousactionsinthein- opment of the theory over the years. By the fall of formationenvironment,incontrasttothenear-certain 1993, Ihadenoughmaterial topresenta seminar at costs and benefits of actions taken in traditional cog- theUniversityofCalifornia—Berkeleycalled‘‘Sense nitivemodelingdomainsofthetime. Making in ComplexInformation Ecologies.’’ It was clear thatit was going to be a challenge to Inthedecadethatfollowed,thefruitfulnessofIn- develop theories for information-intensive tasks. Mul- formationForagingTheorywasapparentfromtheway lingaboutthisissue,Iwasdrawntoworkintwoareasin thatitcouldbeusedtobringmessydataintocrystal whichIhaddonesomereading.Thefirstwasthework clear focus. The first time this happened was in ap- inthelate1980sofJohnR.Anderson(e.g.,Anderson, plication to the Scatter/Gather study presented in 1990),whowasputtingforththeargumentthattoun- chapter 6. Simple analyses of the logs of users inter- derstandmechanismsofthemind,onemustfirsttryto acting with the system seemed to indicate that users figure out the environmental problems that it solves. wherebehavinginanonsystematicwayintheirallo- John developed the method of rational analysis and cationoftimeorintheirchoicesofinterfaceactions. appliedthisapproachtomemory,categorization,and The application of optimal foraging models resulted other areas of cognition with considerable success. I in another of those ‘‘ah-ha’’ experiences in which wondered if the approach could be applied to the suddenly the data plots all fell neatly on lines pre- analysis of the information environment and how it dicted by theory. Like catching a perfect wave in shapesinformationseekingbehavior.1Thesecondarea surfing,thefeelingonegetsfromthatmomentwhen ofinterestwasbehavioralecology(e.g.,Smith,1987), onegainspoweroverasmallportionoftheuniverseis whichsuggestedthatverydiversestrategiesadoptedby hardtorecountwithouttheskillofpoetry(whichIdo peoplecouldbesystematicallypredictedfromoptimi- nothave),anditistherewardthatkeepsyoucoming zationanalysisthatfocusedfirstonscrutinyoftheen- back. vironment.Thisparticularinterestofmineoriginated as an undergraduate at Trent University, where phy- siological psychology included coverage of ethology Acknowledgments (theprecursortobehavioralecology)andanthropology included what is known as cultural materialism (the Inwritingthisbook,Iwasfortunatetohaveinputfrom precursortocurrentevolutionary-ecologicalapproaches agreatpanelofreviewers:MarcMangel,JulieHeiser, to anthropology). Working through the literature in JohnR.Anderson,andJakobNielsen.JacquiLeBlanc theseareas,I was ledtooptimal foraging theory,and read the earliest versions of the manuscript. Each particularlytothebookbyStephensandKrebs(1986) provided a unique perspective from their respective thatisthesourceoftheconventionalmodelsdiscussed fields. I am particularly grateful to Marc, who made inchapter2.Iquiteliterallyhadan‘‘ah-ha’’experience manysuggestionsaboutthemathandaboutconnec- inthemiddleofalate-nightconversationwithJacqui tionstoaricherhistoryofworkin‘‘traditional’’fora- LeBlancinwhichIlaidoutthebasicanalogiesbetween ging theory. I have also been fortunate to work with informationforagingandoptimalforagingtheory. Alex Kirlik, an editor who provided much-needed InJuly1992,Iwroteaworkingpapertitled‘‘Notes collegialadvicethroughoutthisproject. on Adaptive Sense Making in Information Ecolo- PARC has been an especially fertile and sup- gies,’’ which discussed the possible application of portiveenvironmentandImustespeciallythankmy conventional foraging models and the core mathe- managersStuCard,KrisHalvorsen,andMarkStefik matics of Stephens and Krebs to idealized informa- for their continued interest in this work. I also have tionforagingtasks.Theworkingpapergottwokinds been the beneficiary of funding support from the Preface ix most enlightened government funding agencies. I Morrison, Rob Reeder, Pam Schraedley, Mija Van must thank the Office of Naval Research for con- DerWege,andVikramJiswalcontributedenormously tinuedfundingovermanyyearsandintellectualsup- to the efforts to study Web users. Marti Hearst pro- portfromthreegreatprogrammanagers,HelenGig- vided data that proved to be crucial to the analysis ley, Astrid Schmidt-Nielsen, and Susan Chipman. of Scatter/Gather and collaborated with me on the The Advanced Development Research Activity has Scatter/Gather studies along with Patti Schank and provided substantial funding support and sustained ChrisDiehl.IalsothankJanPedersenforinvitingmeto passion and interest from Lucy Nowell and Heather workonScatter/Gatherinthefirstplace.JakobNielsen McMonagle.TheSpencerFoundationprovideddis- hasbeenmyguideinunderstandinghowinformation cretionaryfundingaspartofmyNationalAcademyof foragingtheoryrelatestoconcreteWebusabilityissues. Education Fellowship that was used to support my Overtheyears,JaredSpoolhasdevelopedthenotion dilettante ventures into behavioral ecology and evo- of information scent as a conceptual tool for practi- lutionary-ecologicalanthropology. tioners, and he has always been generous in sharing Manypeoplehavecollaboratedandcontributedto the specific guidelinesthat have beendeveloped asa thisprojectovertheyears.StuCardhelpedmeshape result(manyofthesearepresentedinchapter9). manyoftheideasbyaskingtherightquestions.Ber- Finally, I thank Jacqui LeBlanc for being there nardo Huberman provided a wealth of innovative when the lightning first struck and for her support ideas from a physicist’s perspective, most notably during the first chapter drafts written during an about ultradiffusion, random graph processes, and idyllic stretch of time on the porch of the Dolphin cooperativecomputationalprocesses.JimPitkowwas Inn in Cayucos after morning surf sessions. A more theforcethatgotmanyofusinterestedinanalyzing lovingandlovely museIcouldnot ask for. theemergentdynamicalpropertiesoflargeaggregates of content, users, and topology on the Web. Special thanksgotoDanRussell,whogotallofusinterested Note in sense making in the first place. Pamela E. Sand- stromindependentlydiscoveredtheideaofInforma- 1. As a graduate student working with Anderson, I have notes and working papers from 1982 in which tionForagingTheoryinherworkonscholarlycom- Andersonwasalreadybeginningtosuggestthatfunction munication(chapter8),andshehasgraciouslyshared with respect to the environment would be a crucial to her insights and results over the years. Lada Adamic developingandevaluatingtheoriesofcognition. providedmewithunpublisheddatausedtocalculate thecorrelationbetweeninlinksandoutlinksinchap- ter3.EdChitookupthenotionofinformationscent References andcashedtheideaintorealusabilityanalysissystems, andmorerecentlyhastakenontheideaofsocialin- Anderson, J. R. (1990). The adaptive character of formationforagingasthebackbonefornewinforma- thought. Hillsdale, NJ: Lawrence Erlbaum Associ- tionaccesstechniques.Wai-TatFuhelpedmoveIn- ates. formationForagingTheoryintotherealmoftheWeb Card,S.K.,Mackinlay,J.D.,&Schneiderman,B.(1999). withhisleadershipontheSNIF-ACTproject.Ayman Informationvisualization:Usingvisiontothink.San Farahat,ChristiaanRoyer,andRalucaBudiuhelped Francisco:Morgan-Kaufman. developedahardenedsystemforgeneratingspreading Rao,R.,Pedersen,J.O.,Hearst,M.A.,Mackinlay,J.D., Card, S. K., Masinter, L., et al. (1995). Rich activationnetworksfromonlinecollectionstoreplace interaction in the digital library. Communications the ad hoc code I initially started with. Hinrich oftheACM,38(4),29–39. Schuetzeiscreditedwithprovidingmewiththefirst Smith, E. A. (1987). Optimization theory in anthro- statisticsfromalargedocumentcorpusthatwereused pology: Applications and critiques. In J. Dupre´ todemonstratethatspreadingactivationnetscouldbe (Ed.), The latest on the best (pp. 201–249). Cam- used to predict information scent. Sara Kiesler pro- bridge,MA:MITPress. videdusefulrecommendationsontheliteratureonthe Stephens,D.W.,&Krebs,J.R.(1986).Foragingtheory. relationofcooperativeprocessesandinnovation.Julie Princeton,NJ:PrincetonUniversityPress.
Description:The list of books you might like
