All Chapters Covered
k k
SOLUTION MANUAL
k
,Contents
1 Introduction 3
1.11k Exercises ................................................................................................................................................................ 3
2 Datak Preprocessing 13
2.8 Exercises ...............................................................................................................................................................13
3 Datak Warehousek andk OLAPk Technology:k Ank Overview 31
3.7 Exercises ...............................................................................................................................................................31
4 Datak Cubek Computationk andk Datak Generalization 41
4.5 Exercises ...............................................................................................................................................................41
5 Miningk Frequentk Patterns,k Associations,k andk Correlations 53
5.7 Exercises ...............................................................................................................................................................53
6 Classificationkandk Prediction 69
6.17k Exercises ...............................................................................................................................................................69
7 Clusterk Analysis 79
7.13k Exercises ...............................................................................................................................................................79
8 Miningk Stream,k Time-Series,k andk Sequencek Data 91
8.6 Exercises ...............................................................................................................................................................91
9 Graphk Mining,k Socialk Networkk Analysis,k andk Multirelationalk Datak Mining 103
9.5 Exercises .............................................................................................................................................................103
10 Miningk Object,k Spatial,k Multimedia,k Text,k andk Webk Data 111
10.7 Exercises .............................................................................................................................................................111
11 Applicationsk andk Trendsk ink Datak Mining 123
11.7 Exercises .............................................................................................................................................................123
1
,Chapter 1 k
Introduction
1.11 Exercises
1.1. Whatkiskdatakminingk?k Ink yourkanswer,kaddresskthekfollowing:
(a) Isk itk anotherk hype?
(b) Isk itk ak simplek transformationk ofk technologyk developedk fromk databases,k statistics,k andk machinek learning?
(c) Explaink howk thek evolutionk ofk databasek technologyk ledk tok datak mining.
(d) Describek thek stepsk involvedk ink datak miningk whenk viewedk ask ak processk ofk knowledgek discovery.
Answer:
Datakminingkrefersktokthekprocesskorkmethodkthatkextractskork“mines”kinterestingkknowledgekorkpatternsk
fromklargekamountskofkdata.
(a) Isk itk anotherk hype?
Datakminingkisknotkanotherkhype.k Instead,k thekneedkforkdatakminingkhaskarisenkduektokthekwidek availabilit
ykofkhugekamountskofkdatakandkthekimminentkneedkforkturningksuchkdatakintokusefulkinformationkandkknowl
edge.k Thus,kdatakminingkcankbekviewedkaskthekresultkofktheknaturalkevolutionkofkinformationktechnology.
(b) Iskitkaksimplektransformationkofktechnologykdevelopedkfromkdatabases,kstatistics,kandkmachineklearning?
kNo.k Datakminingkiskmorekthankaksimplektransformation kofktechnology kdeveloped kfromkdatabases,ksta
-
ktistics,k andk machinek learning.k Instead,k datak miningk involvesk ank integration, k ratherk thank ak simpl
e
transformation,k ofk techniquesk fromk multiplek disciplinesk suchk ask databasek technology,k statistics,k ma-
chineklearning,khigh-
performancekcomputing,kpatternkrecognition,kneuralknetworks,kdatakvisualization,kinformationk retrieval,k i
magek andk signalk processing,k andk spatialk datak analysis.
(c) Explaink howk thek evolutionk ofk databasek technologyk ledk tok datak mining.
Databasektechnologykbegankwithkthekdevelopmentkofkdatakcollectionkandkdatabasekcreationkmechanisms
kthatkledktokthekdevelopment kofkeffective kmechanisms kforkdatakmanagement kincludingkdatakstoragekan
dkretrieval,kandkquerykandktransactionkprocessing.kTheklargeknumberkofkdatabaseksystemskofferingkquer
ykandktransactionkprocessingkeventuallykandknaturallykledktokthekneedkforkdatakanalysiskandkunderstandin
g.kHence,kdatakminingkbegankitskdevelopmentkoutkofkthisknecessity.
(d) Describek thek stepsk involvedk ink datak miningk whenk viewedk ask ak processk ofk knowledgek discovery.
Thek stepsk involvedk ink datak miningk whenk viewedk ask ak processk ofk knowledgek discoveryk arek ask follows:
• Datakcleaning,kakprocesskthatkremoveskorktransformsknoisekandkinconsistentkdata
• Datak integration,k wherek multiplek datak sourcesk mayk bek combined
3
, 4 CHAPTERk 1.k k INTRODUCTION
• Datakselection,kwherekdatakrelevantktokthekanalysisktaskkarekretrievedkfromkthekdatabase
• Datak transformation,k wherek datak arek transformedk ork consolidatedk intok formsk appropriatek for
kmining
• Datakmining,kankessentialkprocesskwherekintelligentkandkefficientkmethodskarekappliedkinkorderkto
kextractkpatterns
• Patternk evaluation,k ak processk thatk identifiesk thek trulyk interestingk patternsk representingk knowl-
kedgek basedk onksome kinterestingness kmeasures
• Knowledgek presentation,k wherek visualizationk andk knowledgek representationk techniquesk arek usedk
tokpresentkthekminedkknowledgektokthekuser
1.2. Presentkankexamplekwherekdatakminingkiskcrucialktoktheksuccesskofkakbusiness.k Whatkdatakminingkfunctionsk
doeskthiskbusinesskneed?k Canktheykbekperformedkalternativelykbykdatakquerykprocessingkorksimplekstatisticalkan
alysis?
Answer:
Ak departmentk store,k fork example,k cank usek datak miningk tok assistk withk itsk targetk marketingk mailk campaign.k
Usingkdatakminingkfunctionsksuchkaskassociation,kthekstorekcankusekthekminedkstrongkassociationkrulesktokdeterm
inek whichk productsk boughtk byk onek groupk ofk customers k arek likelyk tok leadk tok thek buyingk ofk certainkotherkp
roducts.k Withkthiskinformation,kthekstorekcankthenkmailkmarketingkmaterialskonlyktokthosekkindskofkcustomers k
whok exhibitk ak highk likelihoodk ofk purchasingk additionalk products.k Datak queryk processingk isk usedkforkdatakork
informationkretrievalkandkdoesknotkhavekthekmeanskforkfindingkassociationkrules.k Similarly,ksimplekstatisticalkan
alysiskcannotkhandleklargekamountskofkdataksuchkaskthosekofkcustomerkrecordskinkakdepartmentk store.
1.3. SupposekyourktaskkaskaksoftwarekengineerkatkBig-
Universitykisktokdesignkakdatakminingksystemktokexaminektheirkuniversitykcoursekdatabase,kwhichkcontains
kthekfollowing kinformation: k thekname,kaddress,kandkstatusk(e.g.,kundergraduate korkgraduate)kofkeachkstuden
t,kthekcoursesktaken,kandktheirkcumulativekgradekpointkaveragek(GPA).kDescribekthekarchitecturekyoukwouldkc
hoose.k Whatkiskthekpurposekofkeachkcomponentkofkthiskarchitecture?
Answer:
Ak datak miningk architecture k thatk cank bek usedk fork thisk applicationk wouldk consistk ofk thek followingk majork c
omponents:
• Akdatabase,kdatakwarehouse,korkotherkinformationkrepository,kwhichkconsistskofktheksetkofkdat
abases,kdatakwarehouses,kspreadsheets,korkotherkkindskofkinformationkrepositorieskcontainingkthekstudentka
ndkcoursekinformation.
• Akdatabasekorkdatakwarehousekserver,kwhichkfetcheskthekrelevantkdatakbasedkonkthekusers’kdatakmi
ningkrequests.
• Akknowledgekbasekthatkcontainskthekdomainkknowledgekusedktokguidektheksearchkorktokevaluatekthekinte
restingnesskofkresultingkpatterns.k Forkexample,kthekknowledgekbasekmaykcontainkconceptkhierarchieskandk
metadatak (e.g.,k describingk datak fromk multiplek heterogeneousk sources).
• Akdatakminingkengine,kwhichkconsistskofkaksetkofkfunctionalkmoduleskforktasksksuchkaskclassification,ka
ssociation,k classification,k clusterk analysis,k andk evolutionk andk deviationk analysis.
• Akpatternkevaluationkmodulekthatkworkskinktandemkwithkthekdatakminingkmoduleskbykemployingkinte
restingnessk measuresk tok helpk focusk thek searchk towardsk interestingk patterns.
• Akgraphicalkuserkinterfacekthatkprovideskthekuserkwithkankinteractivekapproachktokthekdatakminingk
system.
k k
SOLUTION MANUAL
k
,Contents
1 Introduction 3
1.11k Exercises ................................................................................................................................................................ 3
2 Datak Preprocessing 13
2.8 Exercises ...............................................................................................................................................................13
3 Datak Warehousek andk OLAPk Technology:k Ank Overview 31
3.7 Exercises ...............................................................................................................................................................31
4 Datak Cubek Computationk andk Datak Generalization 41
4.5 Exercises ...............................................................................................................................................................41
5 Miningk Frequentk Patterns,k Associations,k andk Correlations 53
5.7 Exercises ...............................................................................................................................................................53
6 Classificationkandk Prediction 69
6.17k Exercises ...............................................................................................................................................................69
7 Clusterk Analysis 79
7.13k Exercises ...............................................................................................................................................................79
8 Miningk Stream,k Time-Series,k andk Sequencek Data 91
8.6 Exercises ...............................................................................................................................................................91
9 Graphk Mining,k Socialk Networkk Analysis,k andk Multirelationalk Datak Mining 103
9.5 Exercises .............................................................................................................................................................103
10 Miningk Object,k Spatial,k Multimedia,k Text,k andk Webk Data 111
10.7 Exercises .............................................................................................................................................................111
11 Applicationsk andk Trendsk ink Datak Mining 123
11.7 Exercises .............................................................................................................................................................123
1
,Chapter 1 k
Introduction
1.11 Exercises
1.1. Whatkiskdatakminingk?k Ink yourkanswer,kaddresskthekfollowing:
(a) Isk itk anotherk hype?
(b) Isk itk ak simplek transformationk ofk technologyk developedk fromk databases,k statistics,k andk machinek learning?
(c) Explaink howk thek evolutionk ofk databasek technologyk ledk tok datak mining.
(d) Describek thek stepsk involvedk ink datak miningk whenk viewedk ask ak processk ofk knowledgek discovery.
Answer:
Datakminingkrefersktokthekprocesskorkmethodkthatkextractskork“mines”kinterestingkknowledgekorkpatternsk
fromklargekamountskofkdata.
(a) Isk itk anotherk hype?
Datakminingkisknotkanotherkhype.k Instead,k thekneedkforkdatakminingkhaskarisenkduektokthekwidek availabilit
ykofkhugekamountskofkdatakandkthekimminentkneedkforkturningksuchkdatakintokusefulkinformationkandkknowl
edge.k Thus,kdatakminingkcankbekviewedkaskthekresultkofktheknaturalkevolutionkofkinformationktechnology.
(b) Iskitkaksimplektransformationkofktechnologykdevelopedkfromkdatabases,kstatistics,kandkmachineklearning?
kNo.k Datakminingkiskmorekthankaksimplektransformation kofktechnology kdeveloped kfromkdatabases,ksta
-
ktistics,k andk machinek learning.k Instead,k datak miningk involvesk ank integration, k ratherk thank ak simpl
e
transformation,k ofk techniquesk fromk multiplek disciplinesk suchk ask databasek technology,k statistics,k ma-
chineklearning,khigh-
performancekcomputing,kpatternkrecognition,kneuralknetworks,kdatakvisualization,kinformationk retrieval,k i
magek andk signalk processing,k andk spatialk datak analysis.
(c) Explaink howk thek evolutionk ofk databasek technologyk ledk tok datak mining.
Databasektechnologykbegankwithkthekdevelopmentkofkdatakcollectionkandkdatabasekcreationkmechanisms
kthatkledktokthekdevelopment kofkeffective kmechanisms kforkdatakmanagement kincludingkdatakstoragekan
dkretrieval,kandkquerykandktransactionkprocessing.kTheklargeknumberkofkdatabaseksystemskofferingkquer
ykandktransactionkprocessingkeventuallykandknaturallykledktokthekneedkforkdatakanalysiskandkunderstandin
g.kHence,kdatakminingkbegankitskdevelopmentkoutkofkthisknecessity.
(d) Describek thek stepsk involvedk ink datak miningk whenk viewedk ask ak processk ofk knowledgek discovery.
Thek stepsk involvedk ink datak miningk whenk viewedk ask ak processk ofk knowledgek discoveryk arek ask follows:
• Datakcleaning,kakprocesskthatkremoveskorktransformsknoisekandkinconsistentkdata
• Datak integration,k wherek multiplek datak sourcesk mayk bek combined
3
, 4 CHAPTERk 1.k k INTRODUCTION
• Datakselection,kwherekdatakrelevantktokthekanalysisktaskkarekretrievedkfromkthekdatabase
• Datak transformation,k wherek datak arek transformedk ork consolidatedk intok formsk appropriatek for
kmining
• Datakmining,kankessentialkprocesskwherekintelligentkandkefficientkmethodskarekappliedkinkorderkto
kextractkpatterns
• Patternk evaluation,k ak processk thatk identifiesk thek trulyk interestingk patternsk representingk knowl-
kedgek basedk onksome kinterestingness kmeasures
• Knowledgek presentation,k wherek visualizationk andk knowledgek representationk techniquesk arek usedk
tokpresentkthekminedkknowledgektokthekuser
1.2. Presentkankexamplekwherekdatakminingkiskcrucialktoktheksuccesskofkakbusiness.k Whatkdatakminingkfunctionsk
doeskthiskbusinesskneed?k Canktheykbekperformedkalternativelykbykdatakquerykprocessingkorksimplekstatisticalkan
alysis?
Answer:
Ak departmentk store,k fork example,k cank usek datak miningk tok assistk withk itsk targetk marketingk mailk campaign.k
Usingkdatakminingkfunctionsksuchkaskassociation,kthekstorekcankusekthekminedkstrongkassociationkrulesktokdeterm
inek whichk productsk boughtk byk onek groupk ofk customers k arek likelyk tok leadk tok thek buyingk ofk certainkotherkp
roducts.k Withkthiskinformation,kthekstorekcankthenkmailkmarketingkmaterialskonlyktokthosekkindskofkcustomers k
whok exhibitk ak highk likelihoodk ofk purchasingk additionalk products.k Datak queryk processingk isk usedkforkdatakork
informationkretrievalkandkdoesknotkhavekthekmeanskforkfindingkassociationkrules.k Similarly,ksimplekstatisticalkan
alysiskcannotkhandleklargekamountskofkdataksuchkaskthosekofkcustomerkrecordskinkakdepartmentk store.
1.3. SupposekyourktaskkaskaksoftwarekengineerkatkBig-
Universitykisktokdesignkakdatakminingksystemktokexaminektheirkuniversitykcoursekdatabase,kwhichkcontains
kthekfollowing kinformation: k thekname,kaddress,kandkstatusk(e.g.,kundergraduate korkgraduate)kofkeachkstuden
t,kthekcoursesktaken,kandktheirkcumulativekgradekpointkaveragek(GPA).kDescribekthekarchitecturekyoukwouldkc
hoose.k Whatkiskthekpurposekofkeachkcomponentkofkthiskarchitecture?
Answer:
Ak datak miningk architecture k thatk cank bek usedk fork thisk applicationk wouldk consistk ofk thek followingk majork c
omponents:
• Akdatabase,kdatakwarehouse,korkotherkinformationkrepository,kwhichkconsistskofktheksetkofkdat
abases,kdatakwarehouses,kspreadsheets,korkotherkkindskofkinformationkrepositorieskcontainingkthekstudentka
ndkcoursekinformation.
• Akdatabasekorkdatakwarehousekserver,kwhichkfetcheskthekrelevantkdatakbasedkonkthekusers’kdatakmi
ningkrequests.
• Akknowledgekbasekthatkcontainskthekdomainkknowledgekusedktokguidektheksearchkorktokevaluatekthekinte
restingnesskofkresultingkpatterns.k Forkexample,kthekknowledgekbasekmaykcontainkconceptkhierarchieskandk
metadatak (e.g.,k describingk datak fromk multiplek heterogeneousk sources).
• Akdatakminingkengine,kwhichkconsistskofkaksetkofkfunctionalkmoduleskforktasksksuchkaskclassification,ka
ssociation,k classification,k clusterk analysis,k andk evolutionk andk deviationk analysis.
• Akpatternkevaluationkmodulekthatkworkskinktandemkwithkthekdatakminingkmoduleskbykemployingkinte
restingnessk measuresk tok helpk focusk thek searchk towardsk interestingk patterns.
• Akgraphicalkuserkinterfacekthatkprovideskthekuserkwithkankinteractivekapproachktokthekdatakminingk
system.
