Mining Behavior of Citizen Sensor Communities to Improve Cooperation with Organizational Actors

155 Pages·2016·6.85 MB·English

by  Hemant Purohit

WWrriigghhtt SSttaattee UUnniivveerrssiittyy
CCOORREE SScchhoollaarr
The Ohio Center of Excellence in Knowledge-
Kno.e.sis Publications
Enabled Computing (Kno.e.sis)
2015
MMiinniinngg BBeehhaavviioorr ooff CCiittiizzeenn SSeennssoorr CCoommmmuunniittiieess ttoo IImmpprroovvee
CCooooppeerraattiioonn wwiitthh OOrrggaanniizzaattiioonnaall AAccttoorrss
Hemant Purohit
Wright State University - Main Campus, [email protected]
Follow this and additional works at: https://corescholar.libraries.wright.edu/knoesis
Part of the Bioinformatics Commons, Communication Technology and New Media Commons,
Databases and Information Systems Commons, OS and Networks Commons, and the Science and
Technology Studies Commons
RReeppoossiittoorryy CCiittaattiioonn
Purohit, H. (2015). Mining Behavior of Citizen Sensor Communities to Improve Cooperation with
Organizational Actors. .
https://corescholar.libraries.wright.edu/knoesis/1082
This Dissertation is brought to you for free and open access by the The Ohio Center of Excellence in Knowledge-
Enabled Computing (Kno.e.sis) at CORE Scholar. It has been accepted for inclusion in Kno.e.sis Publications by an
authorized administrator of CORE Scholar. For more information, please contact [email protected].
Mining Behavior of Citizen Sensor Communities to
Improve Cooperation with Organizational Actors
A dissertation submitted in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
by
Hemant Purohit
B.Tech., The LNM Institute of Information Technology, India, 2009
2015
Wright State University
WrightStateUniversity
GRADUATESCHOOL
August31,2015
I HEREBY RECOMMEND THAT THE DISSERTATION PREPARED UNDER MY
SUPERVISION BY Hemant Purohit ENTITLED Mining Behavior of Citizen Sensor
Communities to Improve Cooperation with Organizational Actors BE ACCEPTED IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF Doctor
ofPhilosophy.
AmitP.Sheth,Ph.D.
DissertationDirector
ArthurA.Goshtasby,Ph.D.
Director,ComputerScience&Engineering,Ph.D.Program
RobertE.W.Fyffe,Ph.D.
VicePresidentforResearch&DeanoftheGraduateSchool
Committeeon
FinalExamination
GuozhuDong,Ph.D.
PatrickMeier,Ph.D.
SrinivasanParthasarathy,Ph.D.
ValerieL.Shalin,Ph.D.
KrishnaprasadThirunarayan,Ph.D.
©Copyrightby
HemantPurohit
2015
ii
ABSTRACT
Purohit, Hemant. Ph.D., Department of Computer Science and Engineering, Wright State Univer-
sity, 2015. MiningBehaviorofCitizenSensorCommunitiestoImproveCooperationwithOrgani-
zationalActors.
Web 2.0 (social media) provides a natural platform for dynamic emergence of citizen (as)
sensor communities, where the citizens generate content for sharing information and en-
gaging in discussions. Such a citizen sensor community (CSC) has stated or implied goals
that are helpful in the work of formal organizations, such as an emergency management
unit, for prioritizing their response needs. This research addresses questions related to
design of a cooperative system of organizations and citizens in CSC. Prior research by so-
cial scientists in a limited offline and online environment has provided a foundation for
research on cooperative behavior challenges, including ‘articulation’ and ‘awareness’, but
Web 2.0 supported CSC offers new challenges as well as opportunities. A CSC presents
information overload for the organizational actors, especially in finding reliable informa-
tionproviders(forawareness),andfindingactionableinformationfromthedatagenerated
by citizens (for articulation). Also, we note three data level challenges–ambiguity in in-
terpreting unconstrained natural language text, sparsity of user behaviors, and diversity of
user demographics. Interdisciplinary research involving social and computer sciences is
essentialtoaddressthesesocio-technicalissues.
I present a novel web information-processing framework, called the Identify-Match-
Engage(IME)framework. IMEallowsoperationalizingcomputationindesignproblemsof
awareness and articulation of the cooperative system between citizens and organizations,
by addressing data problems of group engagement modeling and intent mining. The IME
framework includes: a.) Identification of cooperation-assistive intent (seeking-offering)
fromshort,unstructuredmessagesusingaclassificationmodelwithdeclarative,socialand
contrastpatternknowledge,b.) Facilitationofcoordinationmodelingusingbipartitematch-
ing of complementary intent (seeking-offering), and c.) Identification of user groups to
iii
prioritize for engagement by defining a content-driven measure of group discussion diver-
gence.
The use of prior knowledge and interplay of features of users, content, and network
structuresefficientlycapturescontextforcomputingcooperation-assistivebehavior(intent
and engagement) from unstructured social data in the online socio-technical systems. Our
evaluation of a use-case of the crisis response domain shows improvement in performance
for both intent classification and group engagement prioritization. Real world applica-
tionsofthisworkincludeuseoftheengagementinterfacetoolduringvariousrecentcrises
including the 2014 Jammu and Kashmir floods, and intent classification as a service inte-
gratedbythecrisismappingpioneerUshahidi’sCrisisNETprojectforbroaderimpact.
iv
List of Definitions
• Formal Organization. An organization or institution that has a defined structure of
communication,roles,andwork,e.g.,cityemergencymanagementunit(EMU).
• Organizational actor. A member of the formal organization who understands and
actsfortheorganizationaltasks,processesandworkflows,e.g.,firstresponders.
• Citizen sensor. A user of social media platform, who participates in discussions on
topics related to real world events by generating and sharing information. Roles of
citizensensorsandorganizationalactorsareassumedmutuallyexclusive.
• Citizen sensor community (CSC). A group of citizen sensors on social media who
participateindiscussingvarioustopics. NopriorstructureisassumedinaCSC.
• Goal-orientedCSC.AtypeofCSCwhereusershavevariousintentstoserveagoal,
e.g., a voluntary group during crisis response, a group discussing insights on brand
features,etc.
• Crisis. An escalated emergency event that may be specific, unexpected, and non-
routine event or a series of events. It creates high levels of uncertainty and threat to
anorganization’shighprioritygoalsanditscapacity.
• Behavior. A response to a stimulus environment, e.g., acts of offering help in a
crisis.
• Intent. An aim/plan for (future) action, e.g., wish to donate clothes for help in a
crisis.
• Engagement. A degree of involvement in discussions of a CSC, by participation in
generatingandsharinginformation.
• Coordination. Managingdependenciesbetweentasksinanorganizationalworkflow
bydeliberatejointactions. e.g.,duringcrisis,ateamoforganizationalactorsofEMU
collects information for resource needs from many sources, and processes collected
informationtoachievethegoalofprioritizingresponses.
• Cooperation. Avoluntaryjointactiontohelpotheractorsachievetheirgoal—acon-
trasttocoordination,whichisdeliberateduetomanagingtheinterdependenttasksof
a defined workflow. Cooperation facilitates organizational coordination, e.g., during
crisis, when organizational actors of EMU with defined roles are coordinating (de-
liberately) to collect information on resource needs, CSC members cooperate (vol-
untarily)withthemtohelpminedataonurgentneeds.
• Awareness for Cooperation. A challenge of facilitating shared knowledge among
participatingactorsofcooperation,e.g.,what-where-when-whoduringcrisisresponse.
v
• Articulation for Cooperation. A challenge of managing task divisions and assem-
bling various subtasks and sequences. In order to allow cooperation with citizens,
organizational actors identify information needs specific to their task divisions, e.g.,
duringcrisis,seekingandofferingresourcesarekeyinformationneedsforclearlydi-
vided tasks of resource scarcity and availability information collection, which helps
improvedecisionmakingofprioritizationofresponse.
vi
Contents
1 Chapter1: Introduction 1
1.1 OnlineCitizenSensorCommunity(CSC)andGoal-orientation . . . . . . . 3
1.2 Challenges for Cooperation of Citizens and Organizations: Articulation
andAwareness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 Identify-Match-Engage(IME)FrameworkforaddressingCooperationChal-
lenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.4 IntentMiningandEngagementModelinginIMEFramework . . . . . . . . 8
1.5 ThesisQuestionsandContributions . . . . . . . . . . . . . . . . . . . . . 10
1.6 Use-caseofCrisisResponseDomainandApplications . . . . . . . . . . . 12
1.7 DissertationOrganization . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2 Chapter 2: Verifying Existence of Offline Human Behavior in Online Conver-
sations 16
2.1 InsightsfromOfflineTheoriesofLinguisticCoordination . . . . . . . . . . 17
2.2 TypesofOnlineConversationsFacilitatedbySocialPlatforms . . . . . . . 18
2.2.1 AboutTwitterSocialMedium . . . . . . . . . . . . . . . . . . . . 18
2.2.2 TwitterConversations . . . . . . . . . . . . . . . . . . . . . . . . 19
2.3 Offline Theory-guidedFeatures as Social Knowledgefor Classifying Con-
versations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.4 ClassificationofConversationsinCSC:ExperimentsandResults . . . . . . 25
2.5 Discussion and Hypotheses: Reviewing the Usability of Offline Social
KnowledgeforUnderstandingOnlineSocialData . . . . . . . . . . . . . . 35
3 Chapter 3: Identify function: Intent Classification to Meet Articulation of Or-
ganizationalNeeds 39
3.1 AddressingtheChallengeofMultipleIntentasaClassificationProblem . . 40
3.2 RelatedWorkandtheChallengesofAmbiguityinInterpretation,andSpar-
sityofIntent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.3 Approach v1: Learning with a Bottom-Up Approach of Local Content-
drivenFeatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.4 Approachv2: LearningwithaTop-DownApproachofGlobalKnowledge-
drivenFeatures: Declarative,Social,andContrastPatterns . . . . . . . . . 47
vii
3.5 Approachv3: LearningwithanIntegratedApproachofGlobalKnowledge-
andLocalContent-drivenFeatures . . . . . . . . . . . . . . . . . . . . . . 54
3.6 ExperimentalDesignandImplementation . . . . . . . . . . . . . . . . . . 57
3.7 ResultsandDiscussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
4 Chapter 4: Engage function: User and Group Engagement Modeling for Ad-
dressingAwareness 69
4.1 FindingPrioritizedGroupstoEngagebyModelingDiscussionDivergence . 70
4.2 Relatedwork: ChallengeofDiversityinGroupsofCSC . . . . . . . . . . . 72
4.3 QuantificationofGroupDiscussionDivergence . . . . . . . . . . . . . . . 75
4.4 GroupIdentificationviaCommunityDetectioninInteractionNetwork . . . 76
4.5 Group Representation Features: Quantification of Social Identity and Co-
hesionTheories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.5.1 UserFeatures: Regional,ExpertiseandOnlineIdentities . . . . . . 77
4.5.2 Structural Features: Reciprocity Types in Friendship Network for
ReflectingCohesion . . . . . . . . . . . . . . . . . . . . . . . . . 80
4.6 ExperimentalDesignandImplementation . . . . . . . . . . . . . . . . . . 81
4.6.1 UserandStructuralFeatureCharacteristics . . . . . . . . . . . . . 87
4.6.2 PredictionofTrendforgroupdiscussiondivergence . . . . . . . . 91
4.7 ResultsandDiscussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
5 Chapter5: RealWorldEngagements,Outcomes,andImpact 96
5.1 ApplicationsofIntentandEngagementModelingforaCooperativeSystem 96
5.2 RealworldCrises,andRoleofTechnology: LessonsLearned . . . . . . . . 99
5.3 InterfaceforOrganizationalActorstoCooperatewithCitizens . . . . . . . 100
5.4 IntentClassification-as-a-Service: UshahidiCrisisNETIntegration . . . . . 105
6 Chapter6: Discussion,Limitations,andFutureWork 106
6.1 LessonsonImprovements . . . . . . . . . . . . . . . . . . . . . . . . . . 106
6.1.1 OperationalizingComputationintheCooperativeSystemDesign . 107
6.1.2 DataRepresentationImprovementforIntentandEngagementModels108
6.1.3 FusingTop-downandBottom-UpApproachestoAddressAmbigu-
ity,Sparsity,andDiversity . . . . . . . . . . . . . . . . . . . . . . 108
6.1.4 Importance of Social Behavioral Knowledge in Analyzing Online
SocialData . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6.2 AssumptionsandLimitations . . . . . . . . . . . . . . . . . . . . . . . . . 109
6.2.1 DomainDependence: ContextinCSCWApplications . . . . . . . 109
6.2.2 KnowledgeSources . . . . . . . . . . . . . . . . . . . . . . . . . 110
6.2.3 IntentClasses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
6.2.4 ConsiderationofTemporalDriftintheIntent . . . . . . . . . . . . 110
6.2.5 GroupBehaviorsinEngagementModeling . . . . . . . . . . . . . 111
6.2.6 Non-TwitterSocialData . . . . . . . . . . . . . . . . . . . . . . . 111
6.2.7 InterplayofOfflineandOnlineEnvironments . . . . . . . . . . . . 111
6.2.8 CorrelationbutnotCausalityforAction . . . . . . . . . . . . . . . 111
6.3 FutureWork . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
viii