=Paper=
{{Paper
|id=None
|storemode=property
|title=None
|pdfUrl=https://ceur-ws.org/Vol-811/preface.pdf
|volume=Vol-811
}}
==None==
https://ceur-ws.org/Vol-811/preface.pdf
Proceedingsofthe
RecSys2011
Workshopon
HumanDecisionMakinginRecommender
Systems(Decisions@RecSys’11)
and
UserͲCentricEvaluationofRecommender
SystemsandTheirInterfacesͲ2
(UCERSTI2)
affiliatedwiththe
5thACMConferenceonRecommender
Systems
October23Ͳ27,2011
Chicago,IL,USA
AlexanderFelfernig,LiChen,MonikaMandl,
MartijnWillemsen,DirkBollenandMichaelEkstrand(eds.)
i
Preface(Decisions@RecSys’11)
Interactingwitharecommendersystemmeanstotakedifferentdecisionssuchasselectinga
song/movie from a recommendation list, selecting specific feature values (e.g., camera’s
size, zoom) as criteria, selecting feedback features to be critiqued in a critiquing based
recommendation session, or selecting a repair proposal for inconsistent user preferences
wheninteractingwithaknowledgeͲbasedrecommender.Inallthesescenarios,usershave
to solve a decision task. The major focus of this workshop (Decisions@RecSys’11) are
approachesforefficienthumandecisionmakingindifferentrecommendationscenarios.
Thecomplexityofdecisiontasks,limitedcognitiveresourcesofusers,andthetendencyto
keep the overall decision effort as low as possible leads to the phenomenon of bounded
rationality,i.e.,usersareexploitingdecisionheuristicsratherthantryingtotakeanoptimal
decision.Furthermore,preferencesofuserswilllikelychangethroughoutarecommendation
session,i.e.,preferencesareconstructedinaspecificdecisionenvironmentandusersdonot
knowtheirpreferencesbeforehand.
Decision making under bounded rationality is a door opener for different types of
nonͲconscious influences on the decision behavior of a user. Theories from decision
psychology and cognitive psychology are trying to explain these influences, for example,
decoy effects and defaults can trigger significant shifts in item selection probabilities; in
groupdecisionscenarios,thevisibilityofthepreferencesofothergroupmemberscanhave
asignificantimpactonthefinalgroupdecision.
Themajorgoalofthisworkshopis(was)toestablishaplatformforindustryandacademiato
present and discuss new ideas and research results that are related to human decision
makinginrecommendersystems.Theworkshopconsists(consisted)oftechnicalsessionsin
which results of ongoing research are (were) presented, informal group discussions on
focusedtopics,andakeynotetalkgivenbyAnthonyJamesonfromDFKI,Germany.
Thetopicsofpaperssubmittedtotheworkshopcanbesummarizedasfollows:
x Decision heuristics: the role of decision heuristics/phenomena (e.g., decoys and
anchoring)intheconstructionofrecommenderapplications.
x Recommender user interfaces: impact of recommender interfaces on human decisionͲ
makingbehavior.
x Group decision making: group recommendation algorithms and group decision
strategies.
x EmotionͲbasedrecommendation:emotiondetectionandemotionͲawarerecommender
applications.
x Newapplicationdomains:smarthomesandintelligentdatamanagement.
Theworkshopmaterial(listofacceptedpapers,invitedtalk,andtheworkshopschedule)can
befoundattheDecisions@RecSys2011workshopwebpage:recex.ist.tugraz.at/RecSysWorkshop.
AlexanderFelfernig,LiChen,andMonikaMandl
October2011
ii
Preface(UCERSTI2)
Researchon”HumanͲRecommenderInteraction”isscarce.AlgorithmoptimizationandoffͲ
line testing using measures like RMSE are dominant topics in the RecSys community, but
theorizingaboutconsumerdecisionprocessesandmeasuringusersatisfactioninonlinetests
is less common. Researchers in Marketing and DecisionͲMaking have been investigating
consumer choice processes in great detail, but only sparingly put thisknowledge to use in
technologicalapplications.HumanͲComputerInteractionhasbeenfocusingontheusability
of interfaces for ages, but does not seem to link research on consumer choice and
recommendersysteminterfaces.
During RecSys 2010, we organized the first UCERSTI workshop to bridge these gaps. Two
keynotespeeches,7acceptedpapersandalivelypaneldiscussionintroducedthevisitorsof
RecSys 2010 to the field of HumanͲRecommender Interaction. By means of UCERSTI 2 we
hope to further strengthen the bonds between these researchers, to exchange new
experiences, and meet other new researchers working on userͲcentric research in
RecommenderSystems.
Thepaperscoverthefollowingtopics:
x PreferenceelicitationmethodsandDecisionMakingresearch
x ApplicationsofpsychologicaltheoryandmodelsinRecommendersystems
x UserͲadaptiverecommenderinterfaces
x Quantitativeevaluationofrecommendersystemssuchascontrolledexperimentsand
fieldtrials
x UserͲrecommenderinteractionmeasurementtechniquessuchasquestionnairesand
processdataanalysis
x Useracceptanceofrecommendersystems
UCERSTI 2 also includes a panel discussion, introduced by Joseph A. Konstan and Bart
Knijnenburg,on"Recommendersystemevaluation:creatingaunified,cumulativescience”.
Paneldescription:
Theevaluationofrecommendersystemsistypifiedbyaproliferationofclaims,metricsand
procedures. A review of research papers in Recommender Systems shows a number of
typicalclaims:
x Thisisaninnovativewayofrecommending
x Thisalgorithmismoreaccuratethanothers
x Thisalgorithmisfasterforlargedatasetsthanothers
iii
x This algorithm is better than others along a particular dimension (e.g., diversity,
novelty)
x Thiswayofelicitingratingsleadstogreateraccuracyofrecommendations
x Thisrecommendersystem(algorithm,interface,etc.)ispreferredbyusers
x Thisrecommendersystem(algorithm,interface,etc.)leadstogreaterlongͲtermuser
retentionthanothersystems
For each of these claims recommender systems researchers and practitioners have
developedseveraldistinctmetricstoevaluatethem,aswellasadiversesetofproceduresto
conducttheevaluation.Thisapparentheterogeneitystandsinthewayofscientificprogress.
Researchers face the impossible challenge of selecting a subset of
claims/metrics/proceduresthatallowsforcomparabilityoftheirworkwithpreviousstudies.
Tocreatearigorous,cumulativescienceofrecommendersystems,weneedtotakeastep
backandreflectonourcurrentpractices.
Thisreflectionispartlyphilosophical:Whichofthepossibleinvestigativeclaimsareworthy
of our consideration? The answer to this question depends on the purpose or goal we
ascribe to a recommender system, whom we feel should benefit from it, and where we
believethefieldofrecommendersystemsblendsintootherfields.Inotherwords,weneed
todecideonwhata”goodrecommendersystem”reallyis.
It is also partly practical: As scientists, we need to understand best practices for providing
theevidencetobackuptheseclaims,andforprovidingsuchevidenceinawaythatallows
ourfieldtomoveforward.Someclaims(e.g.,novelty)cansimplybesupportedbyareview
of related work. Others (e.g., user satisfaction) require careful experimental designs that
isolateandmakesalientasmuchaspossiblethefactorbeingstudiedsothatdifferencesin
results can be attributed to that factor. Still others (e.g., algorithmic performance) require
standardization of metrics and evaluation procedures to ensure applesͲtoͲapples
comparisonsagainstthebestpriorwork.
Thispanelwilladdressthegeneralchallengeofbuildingarigorous,cumulativescienceoutof
recommender systems with a specific focus on experiment design and standardization in
supportofbetteruserͲcenteredevaluation.
MoreinformationonUCERSTI2at:http://ucersti.ieis.tue.nl/
MartijnWillemsen,DirkBollenandMichaelEkstrand
October2011
iv
WorkshopCommittee(Decisions@RecSys’11)
Chairs
AlexanderFelfernig,GrazUniversityofTechnology
LiChen,HongKongBaptistUniversity
Organization
MonikaMandl,GrazUniversityofTechnology
ProgramCommittee
MathiasBauer,Mineway,Germany
ShlomoBerkovsky,CSIRO,Australia
RobinBurke,DePaulUniversity,USA
LiChen,HongKongBaptistUniversity,China
HendrikDrachsler,OpenUniversityoftheNetherlands
AlexanderFelfernig,GrazUniversityofTechnology,Austria
GerhardFriedrich,UniversityofKlagenfurt,Austria
SergiuGordea,AustrianInstituteforTechnology,Austria
MehmetGoker,Salesforce,USA
AndreasHolzinger,MedicalUniversityGraz,Austria
DietmarJannach,UniversityofDortmund,Germany
AlfredKobsa,UniversityofCalifornia,USA
GerhardLeitner,UniversityofKlagenfurt,Austria
WalidMaalej,TechnischeUniversitätMünchen,Germany
MonikaMandl,GrazUniversityofTechnology,Austria
FranciscoMartin,BigML,USA
AlexandrosNanopoulos,UniversityofHildesheim,Germany
FrancescoRicci,UniversityofBolzano,Italy
OlgaSantos,UNED,Spain
MonikaSchubert,GrazUniversityofTechnology,Austria
MarkusStrohmaier,GrazUniversityofTechnology,Austria
ErichTeppan,UniversityofKlagenfurt,Austria
NavaTintarev,UniversityofAberdeen,UK
MarcTorrens,Strands,Spain
AlexTuzhilin,NewYorkUniversity,USA
MarkusZanker,UniversityofKlagenfurt,Austria
ChristophZehentner,GrazUniversityofTechnology,Austria
AdditionalReviewers
GeMouzhi,UniversityofDortmund,Germany
v
WorkshopCommittee(UCERSTI)
Organization
MartijnWillemsen,HumanͲTechnologyInteraction,EindhovenUniversityofTechnology,
Netherlands
DirkBollen,HumanͲTechnologyInteraction,EindhovenUniversityofTechnology,
Netherlands
MichaelEkstrand,GroupLensResearch,DepartmentofComputerScienceandEngineering
UniversityofMinnesota,USA
ProgramCommittee
BenedictG.C.Dellaert,DepartmentofBusinessEconomics,ErasmusUniversityRotterdam,
TheNetherlands
MaciejDabrowski,DigitalEnterpriseResearchInstitute,NationalUniversityofIreland,
Galway,Ireland
AlexanderFelfernig,SoftwareTechnologyInstitute,GrazUniversityofTechnology,Germany
DavidGeerts,CentreforUserExperienceResearch,UniversityofLeuven,Belgium
KristiinaKarvonen,HelsinkiInstituteforInformationTechnologyHIIT,Aalto,Finland
AlfredKobsa,DonaldBrenSchoolofInformationandComputerSciences,Universityof
California,Irvine,USA
BartKnijnenburg,DonaldBrenSchoolofInformationandComputerSciences,Universityof
California,Irvine,USA
ArtusKrohnͲGrimberghe,InformationSystemsandMachineLearningLab,Universityof
Hildesheim,Germany
SeanM.McNee,FTITechnology,USA
SteffenRendle,SteffenRendle,SocialNetworkAnalysis,UniversityofKonstanz,Germany
SylvainSenecal,DepartmentofMarketing,HECMontreal,Canada
vi
TableofContents(AcceptedFullPapers)
Decisions@RecSys2011
DecoyEffectsinFinancialServiceEͲSalesSystems
E.Teppan,K.Isak,andA.Felfernig................................................1
Affectiverecommendersystems:theroleofemotionsinrecommendersystems
M.Tkalcic,A.Kosir,andJ.Tasic...................................................9
Usinglatentfeaturesdiversificationtoreducechoicedifficultyinrecommendationlist
M.Willemsen,B.Knijnenburg,M.Graus,L.VelterͲBremmers,andK.Fu................14
Users’DecisionBehaviorinRecommenderInterfaces:ImpactofLayoutDesign
L.ChenandH.Tsoi............................................................21
Visualizableandexplicablerecommendationsobtainedfrompriceestimationfunctions
C.Becerra,F.Gonzalez,andA.Gelbukh..........................................27
RecommenderSystems,ConsumerPreferences,andAnchoringEffects
G.Adomavicius,J.Bockstedt,S.Curley,andJ.Zhang................................35
EvaluatingGroupRecommendationStrategiesinMemoryͲBasedCollaborativeFiltering
N.NajjarandD.Wilson........................................................43
ComputingRecommendationsforLongTermDataAccessibilitybasingonOpenKnowledge
andLinkedData
S.Gordea,A.Lindley,andR.Graf...............................................51
UCERSTI2
AutomatedOntologyEvolutionasaBasisforUserͲAdaptiveRecommenderInterfaces
ElmarP.Wach...............................................................59
AUserͲcentricEvaluationofRecommenderAlgorithmsforanEventRecommendation
System
SimonDooms,ToonDePessemierandLucMartens................................67
EvaluatingRankAccuracybasedonIncompletePairwisePreferences
BrianAckermanandYiChen...................................................74
SettingGoalsandChoosingMetricsforRecommenderSystemEvaluation
GunnarSchroder,MaikThieleandWolfgangLehner................................78
Copyright©2011fortheindividualpapersbythepapers'authors.Copyingpermittedonlyfor
privateandacademicpurposes.Thisvolumeispublishedandcopyrightedbyitseditors:
AlexanderFelfernig,LiChen,MonikaMandl,MartijnWillemsen,DirkBollenandMichaelEkstrand
vii