German clause-embedding predicates [Elektronische Ressource] : an extraction and classification approach / vorgelegt von Ekaterina Lapshinova-Koltunski

Ekaterina Lapshinova-Koltunski

Description

German Clause-EmbeddingPredicates: an Extraction andClassification ApproachVon der Philosophisch-Historischen Fakultät der Universität Stuttgartzur Erlangung der Würde eines Doktors derPhilosophie (Dr. phil.) genehmigte AbhandlungVorgelegt vonEkaterina Lapshinova-Koltunskiaus WolgogradHauptberichter: apl. Prof. Dr. phil. habil. Ulrich HeidMitberichter: Prof. Dr. phil. habil. Grzegorz DogilTag der mündlichen Prüfung: 9. Februar 2010Institut für maschinelle SprachverarbeitungUniversität Stuttgart2009AcknowledgementsThe work on this thesis was performed within the PhD program ’Graduiertenkolleg-609’ funded by DFG. I was supervised by Ulirich Heid at the Institut für MaschinelleSprachverarbeitung of the Universität Stuttgart. I am very grateful to him for hisinterest in my work, his motivation and feedback, and especially for the fact that hecould always find time for me, although it was sometimes difficult due to his full-scale schedule at both universities and his activities in research projects. I would liketo thank him for giving me the opportunity to develop my own ideas: without him Iwould have been just rambling from one topic to another all the time.

Informations

• Publié par: universitat_stuttgart
• Publié le: 01 janvier 2009
• Nombre de lectures: 103
• Poids de l'ouvrage: 1 Mo

Extrait

GermanPredicates:anClause-EmbeddingExtractionand
ApproachClassification

VonderzurErlangungPhilosophisch-HistorischenderWürdeFakultäteinesderDoktorsUniversitätderStuttgart
Philosophie(Dr.phil.)genehmigteAbhandlung

orgelegtVvonoltunskiLapshinova-KEkaterina

Hauptberichter:Mitberichter:

Prüfung:mündlichenderagT

olgogradWaus

apl.Prof.ProfDr..Drphil..phil.habil.habil.GrzegorzUlrichDogilHeid

F9.2010ebruar

InstitutfürUniversitätmaschinelleStuttgartSprachverarbeitung

2009

Acknowledgements

The609’workfundedonbythisDFG.thesisIwaswassupervisedperformedbywithinUlirichtheHeidPhDattheprogramInstitutfür’Graduiertenkolleg-Maschinelle
interestinSprachverarbeitungmywork,ofhisthemotivationUniversitätandStuttgart.feedback,Iandamveryespeciallygratefulforthetofacthimthatforhishe
couldalwaysfindtimeforme,althoughitwassometimesdifficultduetohisfull-
toscalethankschedulehimforatbothgivingmeuniversitiestheandopportunityhistoactivitiesdevelopinmyresearchownideas:projects.IwithoutwouldhimlikeI
wouldhavebeenjustramblingfromonetopictoanotherallthetime.
Iamgratefultoallofthe’Graduiertenkolleg-609’:ourprofessors-Grzegorz
BerndDogil,JürgenMöbius;Pasafel,wellHansasKamp,myfellowKlausvonPhDandHeusingerour,studentsHinrich-Schütze,OlgaAchimAnufryk,Stein,Na-
talieLewandowska,KlausRothenhäusler,DennisSpohr,HenrikeBaumotte,Charles
RYee,egineUtaBennerBrandtner,,SimoneDolgorHeinold,Guntsetseg,ManuelaMelanieKorth,UthPetraandManuelAugurzkyK,untz.ZoricaTTheirrpcevska,help-
fulgungen’commentsinandKleinwalsertal,suggestionsAustria,gaveandmeathelotofentertainingideasduringtrekkingourtoursannualinthe’Klausurta-Alpes
Ibroughtalsousthankallthetogetherasadministrativefriends.staffofthe’Graduiertenkolleg-609’,namely,Eva
suesSchmidIwouldandhaveChristiantogoBär,throughwhohelpedotherwise.metoTheavoidfinancialalotofsupporttheprovidedbureaucraticbytheis-
wholeresearch’Graduiertenkolleg-609’projectmadepossible.myparticipationininternationalconferencesandthis
whoImadewouldsurealsothatliketomythankworkingprofessorsenvironmentandthewasbothadministrativeefficientandstaffatthecomfortable.IMS,
IamsemanticespeciallyissuesandobligedtotoGertrudAntjeFaaß,RSabineoßdeutscherSchulteandKimurtWalde,EberleKlausforRtheirothenhäusleradvicein,
AndréBlessing,JuliaWeidenkaffandFabienneFritzingerfortheirhelpwithtechnical
withproblemscommentsconcerningontheCQPdraftandofthisUNIX.thesis.MyIgratitudeamisgratefulduetotoallthosemywhocolleaguesprovidedattheme
IMSforcreatingafriendlyandstimulatingatmosphere.SpecialthanksgotoOlga
Anufryk,NatalieLewandowska,KlausRothenhäuslerandDennisSpohr,whohelped
meinsidetoorforgetoutsideabouttheworkIMS.forIatalsoleastthankhalfanOlgahourAnufryk,perday,NatalieenjoyingourLewandowskalunchorandcoffeemy
bothofficemates,HannahKermesandOksanaStus,fordiscussingnotonlymatters
concerningFinally,Iworkwouldbutlikealsotoproblemsthankmyconcerningfamilyandsocialmyissues.friendsforalltheirloveand
oversupport.theIworld.amsoOurluckytocommunicationhavealotofhelpedfriendsmeingreatlyR,ussia,beitGermanysometimesand,onlyactuallythrough,all

3

theUlrike,internet.whomadeAnothermyroundalmostofdailyspecialtrainthanksridesgotobetweenmytrainFreiburgmates,andIsabelle,StuttgartAnnfunnyand
andlessstressfull.Iamgratefultoallmyfamilymembers,especiallymygrandma
andmylategrandpa,whowouldbehappytosharesharemyRigorosumwithme.
Ichattedwouldlikealmosttoeverythankmydayaunt,duringunclethelastandmonthscousins,ofmyespeciallyworkoverNatasha,thewithdissertationwhomI
andparents,whoIrinahelpedandmetoSergey,switchmyofffromparents-in-lawthedaily,Irenaroutine.andHugeGerd,asthankswelltoasbothAnnaofandmy
goChristian,abroadforandtheirlivethousandsemotionalofsupport.kilometresSpecialawaythanksfromtothem.myIparentsknowfortheylettingmissmeme
justasmuchasImissthem.And,lastbutnotleast,majorthankstomyhusband
KrzysiekforhispatiencewithmeduringthecountlessdaysthatIspentinfrontof
myeveniflaptopIstoppedinsteaddoingofso.spendingIamthemhappytowithhavehim.himHeonnevermysidestoppedandsharebelievingtheinburdenme
andthefunofmywork.

Contents

1

2

3

andAimsIntroduction:Motivation

1

StateoftheArtonSubcategorisationinLinguisticsandNLP7
2.1TheNotionanditsBasicCategories....................7
2.1.1DevelopmentoftheNotion....................8
2.1.2SyntacticandSemanticLevelsofValency............9
2.1.3Complementation:QuantitativeandQualitativeValency....10
2.1.3.1Quantitativeandqualitativevalency:valencypatterns10
2.1.3.2Complementsandadjuncts...............12
2.1.4ComplementRealisationontheSyntacticLevel.........15
2.1.4.1Grammaticalfunctionsofcomplements........16
2.1.4.2Syntacticcategoriesofcomplements..........19
2.1.5ComplementDescriptionontheSemanticLevel.........20
2.2SytaxandSemanticsofSententialComplements.............22
2.2.1SententialComplementsontheSyntacticLevel.........22
2.2.1.1Formsofsententialcomplements............22
2.2.1.2Positioninasentence..................24
2.2.1.3Grammaticalfunctionsofsubclauses..........25
2.2.2SemanticsofSententialComplements..............27
2.2.2.1Semanticsofdeclarativeclauses............27
2.2.2.2Semanticsofinterrogativeclauses...........28
2.2.2.3Sententialcomplementsvs.predicates:theirsemantics30
2.3Summary:typesofvalencyinformationrelatedtothisstudy......33

SubcategorisationofVerbs,NounsandMultiwords35
3.1SubcategorisationofVerbs.........................36
3.1.1VerbalPredicatesinLexicographicWork.............36
3.1.1.1Verbalpredicatesinprinteddictionaries........36
3.1.1.2Verbalpredicatesinelectronicdictionaries......38
3.1.1.3Summary:verbalpredicatesinlexicography.....40
3.1.2VerbalPredicatesinNLPWork..................40
3.1.2.1Verbalpredicatesinformalgrammars.........40
3.1.2.2VerbalpredicatesinNLP-baseddictionaries......43
3.1.2.3Summary:verbalpredicatesinNLP..........47
3.2SubcategorisationofNouns........................47
3.2.1NominalPredicatesinLinguisticandLexicographicWork....48
3.2.1.1Linguisticstudies.....................48

i

ii

4

CONTENTS

3.2.1.2Nominalpredicatesinlexicography..........49
3.2.1.3Summary:nominalsinlinguisticsandlexicography.52
3.2.2NominalPredicatesinNLPWork.................52
3.2.2.1Nominalpredicatesinformalgrammars........52
3.2.2.2NominalpredicatesinNLP-baseddictionaries.....55
3.2.2.3Summary:nominalsinNLP...............58
3.3SubcategorisationofMultiwordExpressions...............58
3.3.1MultiwordPredicatesinLinguisticandLexicographicWork..59
3.3.1.1Multiwordsinlinguisticstudies.............59
3.3.1.2Multiwordsinlexicography...............61
3.3.2MultiwordPredicatesinNLPWork................62
3.3.3Summary:MultiwordsinLinguisticsandNLP..........65
3.4ThePhenomenonof“Inheritance”ofSubcategorisation.........65
3.4.1“Inheritance”inMultiwordandCompoundPredicates.....66
3.4.1.1Valencyofmultiwords:nominalortheirown?....66
3.4.1.2Valencyofcompoundnouns:headornon-head?...69
3.4.1.3Summary:“inheritance”inmultiwordsandcompounds71
3.4.2“Inheritance”inNominalisations.................72
3.4.2.1Nominalisationsandtheirbaseverbs..........72
3.4.2.2“Inheritance”ofsubcategorisation...........73
3.4.2.3“Non-inheritance”ofsubcategorisation........78
3.4.2.4Reasonsfor“non-inheritance”.............82
3.4.2.5Summary:“inheritance”innominalisations......84
AcquisitionandClassificationofPredicates87
4.1AcquisitionofPredicates..........................87
4.1.1AcquisitionToolsforVerbalPredicates..............87
4.1.1.1VerbalpredicatesinEnglish...............88
4.1.1.2VerbalpredicatesinGermanandDutch........90
4.1.1.3VerbalpredicatesinRomancelanguages:FR,IT,PT.91
4.1.1.4Verbalpredicatesinfurtherlanguages.........92
4.1.2AcquisitionToolsforNominalPredicates.............93
4.1.3AcquisitionToolsforMultiwordPredicates............94
4.1.4Summary:RelatedWorkonPredicateAcquisition........95
4.2ClassificationofPredicates.........................96
4.2.1ClassificationofVerbalPredicates.................96
4.2.1.1Aspectualcriteria....................97
4.2.1.2Semanticandsyntacticcriteria.............98
4.2.1.3Verbclassesrelatedtothisstudy............100
4.2.2ClassificationofNominalPredicates...............101
4.2.2.1Aspectualandfunctionalcriteria............101
4.2.2.2Semanticcriteria.....................102
4.2.2.3Morphologicalandderivationcriteria.........102
4.2.2.4Criteriaaccordingtothesubcategorisationproperties103
4.2.2.5Nominalclassesrelatedtothisstudy..........103
4.2.3ClassificationofCompoundNominals..............104
4.2.3.1Derivationalcriteria...................104

CONTENTS5

iii

4.2.3.2Relationsbetweenheadsandnon-headsascriteria..104
4.2.3.3Classesofcompoundnominalsrelatedtothisstudy.105
4.2.4MultiwordsandtheirClassification................107
4.2.4.1Criteriaofcompositionalityorlexicalisationgrade..107
4.2.4.2Structuralandsyntacticcriteria.............108
4.2.4.3Semanticandaspectualcriteria.............109
4.2.4.4Multiwordclassesrelatedtothisstudy.........110
4.2.5“Inheritance”RelationsandtheirTypes..............111

115ArchitectureClassificationandExtraction5.1Input:CorporaandtheirAnnotation...................115
5.1.1CorporaSpecification.......................115
5.1.2CorpusPre-processingTools....................117
5.1.3ExperimentswithParsedCorpora.................121
5.2ExtractionContext.............................122
5.2.1Contextsforthe