Innovative Approaches in Agent-Based Modelling and Business Intelligence

284 Pages·2018·11.93 MB·English

by  Setsuya Kurahashi

Agent-Based Social Systems 12
Setsuya Kurahashi · Hiroshi Takahashi
Editors
Innovative
Approaches in Agent-
Based Modelling and
Business Intelligence
Agent-Based Social Systems
Volume 12
EditorinChief
HiroshiDeguchi,Yokohama,Japan
SeriesEditors
Shu-HengChen,Taipei,Taiwan,ROC
ClaudioCioffi-Revilla,Fairfax,USA
NigelGilbert,Guildford,UK
HajimeKita,Kyoto,Japan
TakaoTerano,Yokohama,Japan
KyoichiKijima,Tokyo,Japan
SetsuyaKurahashi,Tokyo,Japan
ManabuIchikawa,Saitama,Japan
ShingoTakahashi,Tokyo,Japan
MotonariTanabu,Yokohama,Japan
Aki-HiroSato,Kyoto,Japan
This series is intended to further the creation of the science of agent-based social
systems, a field that is establishing itself as a transdisciplinary and cross-cultural
science. The series will cover a broad spectrum of sciences, such as social sys-
tems theory, sociology, business administration, management information science,
organizationscience,computationalmathematicalorganizationtheory,economics,
evolutionary economics, international political science, jurisprudence, policy sci-
ence, socioinformation studies, cognitive science, artificial intelligence, complex
adaptivesystemstheory,philosophyofscience,andotherrelateddisciplines.
The series will provide a systematic study of the various new cross-cultural
arenas of the human sciences. Such an approach has been successfully tried
several times in the history of the modern science of humanities and systems
and has helped to create such important conceptual frameworks and theories as
cybernetics, synergetics, general systems theory, cognitive science, and complex
adaptivesystems.
Wewanttocreateaconceptualframeworkanddesigntheoryforsocioeconomic
systemsofthetwenty-firstcenturyinacross-culturalandtransdisciplinarycontext.
Forthispurposeweplantotakeanagent-basedapproach.Developedoverthelast
decade, agent-based modeling is a new trend within the social sciences and is a
child of the modern sciences of humanities and systems. In this series the term
“agent-based” is used across a broad spectrum that includes not only the classical
usage of the normative and rational agent but also an interpretive and subjective
agent. We seek the antinomy of the macro and micro, subjective and rational,
functional and structural,bottom-up and top-down, global and local, and structure
and agency within the social sciences. Agent-based modeling includes both sides
oftheseopposites.“Agent”isourgroundingformodeling;simulation,theory,and
realworldgroundingarealsorequired.
As an approach, agent-based simulation is an important tool for the new
experimentalfieldsofthesocialsciences;itcanbeusedtoprovideexplanationsand
decision support for real-world problems, and its theories include both conceptual
andmathematicalones.Aconceptualapproachisvitalforcreatingnewframeworks
of the worldview, and the mathematical approach is essential to clarify the logical
structureofanynewframeworkormodel.Explorationofseveraldifferentwaysof
real-world grounding is required for this approach. Other issues to be considered
intheseriesincludethesystemsdesignofthiscentury’sglobalandlocalsocioeco-
nomicsystems.
Moreinformationaboutthisseriesathttp://www.springer.com/series/7188
Setsuya Kurahashi • Hiroshi Takahashi
Editors
Innovative Approaches in
Agent-Based Modelling and
Business Intelligence
123
Editors
SetsuyaKurahashi HiroshiTakahashi
GraduateSchoolofBusinessSciences KeioUniversity
UniversityofTsukuba Yokohama,Kanagawa,Japan
Tokyo,Tokyo,Japan
ISSN1861-0803
Agent-BasedSocialSystems
ISBN978-981-13-1848-1 ISBN978-981-13-1849-8 (eBook)
https://doi.org/10.1007/978-981-13-1849-8
LibraryofCongressControlNumber:2018963745
©SpringerNatureSingaporePteLtd.2018
Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpartof
thematerialisconcerned,specificallytherightsoftranslation,reprinting,reuseofillustrations,recitation,
broadcasting,reproductiononmicrofilmsorinanyotherphysicalway,andtransmissionorinformation
storageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilarmethodology
nowknownorhereafterdeveloped.
Theuseofgeneraldescriptivenames,registerednames,trademarks,servicemarks,etc.inthispublication
doesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevant
protectivelawsandregulationsandthereforefreeforgeneraluse.
Thepublisher,theauthorsandtheeditorsaresafetoassumethattheadviceandinformationinthisbook
arebelievedtobetrueandaccurateatthedateofpublication.Neitherthepublishernortheauthorsor
theeditorsgiveawarranty,expressorimplied,withrespecttothematerialcontainedhereinorforany
errorsoromissionsthatmayhavebeenmade.Thepublisherremainsneutralwithregardtojurisdictional
claimsinpublishedmapsandinstitutionalaffiliations.
ThisSpringerimprintispublishedbytheregisteredcompanySpringerNatureSingaporePteLtd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Preface
The purpose of this book is to thoroughly prepare the reader at an intermediate
level for research in social science, organization studies, economics, finance,
marketing science, and business science as complex adaptive systems. Those who
are not familiar with a computational research approach may see the advantages
of social simulation studies and business intelligence, and experienced modelers
mayalsofindvariousinstructiveexamplesusingagent-basedmodelingandbusiness
intelligence approaches to inspire their own work. In addition, the book discusses
cutting-edgetechniquesforcomplexadaptivesystemsthroughtheirapplications.
Business science studies so far have focused only on data science and analyses
ofbusinessproblems.However,thesestudiestoenhancethecapabilitiesofconven-
tionaltechniquesinthefieldshavenotbeeninvestigatedadequately.Theemphasis
in this book is managing the issues of societies, firms, and organizations for
achieving profit on interaction with agent-based modeling, human- and computer-
mixed systems, and business intelligence approaches, as such a focus is also
fundamentalforcomplexbutboundedrationalbusinessenvironments.
Appropriate for a diverse readership, there are multiple ways to read this book
depending on readers’ interests in the areas of application and on their level of
technicalskills.Researchersfamiliarwithfieldssuchassocialscienceandbusiness
sciencestudiesmaycomparetheideasexpressedhereusingsimulationmodelswith
empiricalstudies.
Innovative Approaches in Agent-Based Modelling and Business Intelligence
encourages readers inspired by intensive research works by leading authors in the
field to join with other disciplines and extend the scope of the book with their
own unique contributions. Agent-based modeling and business intelligence with
the latest results in this book allow readers who are researchers, students, and
professionals to resolve their problems through the common challenges posed by
computational social and business science researchers involved in both areas in
ordertocreateavaluablesynergy.
Thisbookcontains19chapters.Thefollowingaretheirsynopses.TakaoTerano
discusses the basic principles and key ideas of EBMS which is a just started new
fieldofbothscientificandpracticalactivitiesinhisChap.1.
v
vi Preface
KotaroOhoriandHirokazuAnaipresentanovelresearchproject,whichtheycall
“social mathematics,” to resolve social issues based on mathematical and artificial
intelligencetechnologiesintheirChap.2.
Chathura Rajapakse, Lakshika Amarasinghe, and Kushan Ratnayake present
the details of an agent-based simulation model developed to study the impact of
seepage behavior, which means the smaller vehicles moving forward through the
gapsbetweenlargervehicleswithoutfollowingthelanesinthetrafficcongestionin
theirChap.3.
HiroshiTakahashidiscussestheinfluenceofinformationtechnologyonbusiness
andfinance.Healsolookstoofferanoverviewofthekindofresearchagent-based
modelsareenablinginhisChap.4.
Takashi Ishikawa investigates the mechanism of coevolving networks using a
generalizedadaptivevotermodelbasedonrelatedworkandthehomophilyprinciple
which is known as a driving mechanism to form community structure in social
networksinhisChap.5.
YokoIshinoproposesanewmethodforobtaininganappropriatestructurefora
BayesiannetworkbyusingsensitivityanalysisinastepwisefashioninherChap.6.
Hiroaki Jotaki, Yasuo Yamashita, Satoru Takahashi, and Hiroshi Takahashi
analyzestheinfluenceoftextinformationoncreditmarketsinJapanintheirChap.7.
Itfocusesonheadlinenews,asourceofinformationthathasimmediateinfluenceon
themoneymarketandalsowhichisregardedasanimportantsourceofinformation
whenmakinginvestmentdecisions.
Yasuo Kadono tries to develop an integrated approach of data obtained from
issue-orientedlarge-scalefact-findingsurveys,statisticalanalysesbasedondynamic
modeling,andsimulationsinhisChap.8.
Hajime Kita shows the experience of 20-year study of the artificial market and
discussesitsfutureinhisChap.9.
Masaaki Kunigami introduces a new formulation called the Doubly Structural
Network(DSN)Modelandshowsitsapplicationsinsocioeconomicsandeducation
inhisChap.10.
SetsuyaKurahashilooksbackatthehistoryofscienceinordertointroduceone
method so that ABS can develop to more reliable social science in his Chap.11.
It also overviews the validity of modeling, ABM as an inductive inference and
deductiveinference.
ZhenxiChenandThomasLuxapplythesimulatedmethodofmomentestimator
proposedbyChenandLuxtoinvestigatetheherdingbehaviorintheChinesestock
marketusinghigh-frequencydataintheirChap.12.
AkiraOta,GadeaUriel,HiroshiTakahashi,andToshiyukiKanedadiscussfactors
ofthepedestrianflowsintwodifferentundergroundmallsbyconductingmultiple
regressionanalysiswithvisibilitymeasuresbasedonspacesyntaxtheoryandstore
proximitymeasuresintheirChap.13.
Makoto Sonohara, Kohei Sakai, Masakazu Takahshi, and Toshiyuki Kaneda
focus on tourists’ evacuation behavior who don’t have enough knowledge of
evacuationsitesandroutesintheirChap.14.Theirstudyshowsanagentmodeling
techniqueusingasamplingsurveyoftheweb-basedquestionnaireconsideringthe
Preface vii
informationbehaviorsandtheearthquakeexperiencesandatouristevacuationagent
modelusingthistechnique.
Shingo Takahashi proposes “virtual grounding” as a grounding method for
constructingvalidfacsimilemodelswhererealdataforbehavioralmodelparameter
identificationarenotavailableinhisChap.15.
Toru B. Takahashi proposes a problem-solving support agent that interactively
supportshumanproblemsolvingactivitiesinhisChap.16.Priortothedevelopment
oftheproblem-solvingsupportagent,heorganizedtheproblemsolvingprocessand
researchedthetypesofmistakesinvolvedintheprocess.
Wander Jager, Geeske Scholz, Ren´e Mellema, and Setsuya Kurahashi discuss
theirexperienceswiththeEnergyTransitionGame(ETG)inGroningen,Tokyo,and
Osnabrück,allineducationalsettingsintheirChap.17.TheETGisanagent-based
gameinwhichrolesthatcanbeplayedareenergycompaniesandpoliticalparties.
Auniqueaspectistheinclusionofanartificialpopulationofsimulatedpeople.
ChaoYangproposesaco-evolutionaryopinionmodelofthesocialnetworkbased
onboundedconfidence,referencerange,andinteractiveinfluenceinherChap.18.
Finally,TakashiYamadareviewstheactivitiesofProf.DrTakaoTerano’slabo-
ratoryatTokyoInstituteofTechnologyandbrieflyintroducesseveralrepresentative
papersespeciallyinsocialsimulationliteratureinhisChap.19.
Acknowledgements Astheeditors,wewouldliketothankProf.TakaoTerano.Hehasbeena
leadingresearcherinagent-basedmodelingfieldsforseveraldecades.Finally,wewishtoexpress
ourgratitudetoalltheauthors.
Tokyo,Japan SetsuyaKurahashi
Yokohama,Japan HiroshiTakahashi
June2018
Contents
1 Gallery for Evolutionary Computation and Artificial
IntelligenceResearches:WhereDoWeComefromandWhere
ShallWeGo ................................................................. 1
TakaoTerano
2 MathematicalTechnologiesandArtificialIntelligenceToward
Human-CentricInnovation................................................ 9
KotaroOhoriandHirokazuAnai
3 Study on the Social Perspectives of Traffic Congestion
inSriLankaThroughAgent-BasedModelingandSimulation:
LessonsLearnedandFutureProspects.................................. 23
ChathuraRajapakse,LakshikaAmarasinghe,andKushanRatnayake
4 InformationTechnologyandFinance .................................... 43
HiroshiTakahashi
5 TwoPhaseTransitionsintheAdaptiveVoterModelBased
ontheHomophilyPrinciple ............................................... 53
TakashiIshikawa
6 SensitivityAnalysisinaBayesianNetworkforModelinganAgent .. 65
YokoIshino
7 AnalyzingtheInfluenceofHeadlineNewsonCreditMarkets
inJapan...................................................................... 77
Hiroaki Jotaki, Yasuo Yamashita, Satoru Takahashi,
andHiroshiTakahashi
8 ConsiderationonanIntegratedApproachtoSolvingIndustrial
IssuesThroughSurveys,Statistics,andSimulations ................... 95
YasuoKadono
9 U-Mart:20-YearExperienceofanArtificialMarketStudy........... 111
HajimeKita
ix
x Contents
10 What Do Agents Recognize? From Social Dynamics
toEducationalExperiments............................................... 123
MasaakiKunigami
11 ModelPredictionandInverseSimulation ............................... 139
SetsuyaKurahashi
12 IdentificationofHigh-Frequency HerdingBehaviorinthe
ChineseStockMarket:AnAgent-BasedApproach .................... 157
ZhenxiChenandThomasLux
13 ADataAnalysisStudyonFactorsofthePedestrianFlows
in Two Different Underground Malls Using Space Syntax
Measures:CaseComparisonsinNagoya,Japan........................ 173
AkiraOta,GadeaUriel,HiroshiTakahashi,andToshiyukiKaneda
14 AStudyonAgentModelingofTouristEvacuationBehaviors
inanEarthquake:ACaseStudyofanEvacuationSimulation
ofHimejiCastle............................................................. 189
Makoto Sonohara, Kohei Sakai, Masakazu Takahshi,
andToshiyukiKaneda
15 VirtualGroundingforAgent-BasedModelinginIncomplete
DataSituation............................................................... 205
ShingoTakahashi
16 AnalysisofProblem-SolvingProcesses................................... 221
ToruB.Takahashi
17 TheEnergyTransitionGame:ExperiencesandWaysForward...... 237
WanderJager,GeeskeScholz,Ren´eMellema,andSetsuyaKurahashi
18 A Coevolutionary Opinion Model Based on Bounded
Confidence,ReferenceRange,andInteractiveInfluencein
SocialNetwork.............................................................. 253
ChaoYang
19 Prof.Dr.TakaoTeranoasaBrilliantEducator......................... 269
TakashiYamada