计算语言学协会会刊, 卷. 6, PP. 197–210, 2018. 动作编辑器: Hinrich Sch¨utze. 提交批次: 6/2017; 修改批次: 9/2017; 已发表 4/2018. 2018 计算语言学协会. 根据 CC-BY 分发 4.0 执照. C (西德:13) KnowledgeCompletionforGenericsusingGuidedTensorFactorizationHanieSedghi∗GoogleBrainMountainView,CA,U.S.A.hsedghi@google.comAshishSabharwalAllenInstituteforArtificialIntelligence(AI2)Seattle,WA,U.S.A.AshishS@allenai.orgAbstractGivenaknowledgebaseorKBcontaining(嘈杂)factsaboutcommonnounsorgener-ics,suchas“alltreesproduceoxygen”or“someanimalsliveinforests”,weconsidertheproblemofinferringadditionalsuchfactsataprecisionsimilartothatofthestartingKB.SuchKBscapturegeneralknowledgeabouttheworld,andarecrucialforvariousappli-cationssuchasquestionanswering.Differ-entfromcommonlystudiednamedentityKBssuchasFreebase,genericsKBsinvolvequan-tification,havemorecomplexunderlyingreg-ularities,tendtobemoreincomplete,andvio-latethecommonlyusedlocallyclosedworldassumption(LCWA).WeshowthatexistingKBcompletionmethodsstrugglewiththisnewtask,andpresentthefirstapproachthatissuccessful.Ourresultsdemonstratethatex-ternalinformation,suchasrelationschemasandentitytaxonomies,ifusedappropriately,canbeasurprisinglypowerfultoolinthisset-ting.First,oursimpleyeteffectiveknowledgeguidedtensorfactorizationapproachachievesstate-of-the-artresultsontwogenericsKBs(80%precise)forscience,doublingtheirsizeat74%-86%precision.Second,ournoveltax-onomyguided,submodular,activelearningmethodforcollectingannotationsaboutrareentities(e.g.,oriole,abird)is6xmoreeffec-tiveatinferringfurthernewfactsaboutthemthanmultipleactivelearningbaselines.1IntroductionWeconsidertheproblemofcompletingapartialknowledgebase(KB)containingfactsaboutgener-∗ThisworkwasdonewhiletheauthorwasaffiliatedwiththeAllenInstituteforArtificialIntelligence.icsorcommonnouns,representedasathird-ordertensorof(来源,关系,目标)三元组,suchas(butterfly,pollinate,flower)和(thermometer,mea-sure,temperature).Suchfactscapturecommonknowledgethathumanshaveabouttheworld.Theyarearguablyessentialforintelligentagentswithhuman-likeconversationalabilitiesaswellasforspecificapplicationssuchasquestionanswering.Wedemonstratethatstate-of-the-artKBcompletionmethodsperformpoorlywhenfacedwithgener-ics,whileourstrategiesforincorporatingexternalknowledgeaswellasobtainingadditionalannota-tionsforrareentitiesprovidethefirstsuccessfulso-lutiontothischallengingnewtask.Sincegenericsrepresentclassesofsimilarindi-viduals,thetruthvalueyiofagenericstriplexi=(s,r,t)dependsonthequantificationsemanticsoneassociateswithsandt.Indeed,thesemanticsofgenericsstatementscanbeambiguous,evenself-contradictory,duetoculturalnorms.AsLeslie(2008)pointsout,‘duckslayeggs’isgenerallycon-sideredtruewhile‘ducksarefemale’,whichistrueforabroadersetofducksthantheformerstatement,isgenerallyconsideredfalse.Toavoiddeepphilosophicalissues,wefixapar-ticularmathematicalsemanticsthatisespeciallyrel-evantfornoisyfactsderivedautomaticallyfromtext:associateswithacategoricalquantificationfrom{全部,一些,none}andassociatet(implicitly)withsome.Forinstance,“allbutterfliespollinate(一些)flower”and“someanimalslivein(一些)forest”.Whenpresentingsuchtriplestohumans,theyarephrasedas:isittruethatallbutterfliespollinatesomeflower?Asanotationalshortcut,wetreatthequantificationofsasthecategoricallabelyiforthetriplexi.Forexample,(butterfly,pollinate,flower) l D o w n o a d e d f r o m h t t p : / / 直接的

计算语言学协会会刊, 卷. 6, PP. 159–172, 2018. 动作编辑器: Luke Zettlemoyer. 提交批次: 10/2017; 修改批次: 12/2017; 已发表 3/2018. 2018 计算语言学协会. 根据 CC-BY 分发 4.0 执照. C (西德:13) MappingtoDeclarativeKnowledgeforWordProblemSolvingSubhroRoy∗MassachusettsInstituteofTechnologysubhro@csail.mit.eduDanRoth∗UniversityofPennsylvaniadanroth@seas.upenn.eduAbstractMathwordproblemsformanaturalabstrac-tiontoarangeofquantitativereasoningprob-lems,suchasunderstandingfinancialnews,sportsresults,andcasualtiesofwar.Solvingsuchproblemsrequirestheunderstandingofseveralmathematicalconceptssuchasdimen-sionalanalysis,subsetrelationships,etc.Inthispaper,wedevelopdeclarativeruleswhichgovernthetranslationofnaturallanguagede-scriptionoftheseconceptstomathexpres-sions.Wethenpresentaframeworkforin-corporatingsuchdeclarativeknowledgeintowordproblemsolving.Ourmethodlearnstomaparithmeticwordproblemtexttomathex-pressions,bylearningtoselecttherelevantdeclarativeknowledgeforeachoperationofthesolutionexpression.Thisprovidesawaytohandlemultipleconceptsinthesameprob-lemwhile,atthesametime,supportingin-terpretabilityoftheanswerexpression.Ourmethodmodelsthemappingtodeclarativeknowledgeasalatentvariable,thusremov-ingtheneedforexpensiveannotations.Exper-imentalevaluationsuggeststhatourdomainknowledgebasedsolveroutperformsallothersystems,andthatitgeneralizesbetterintherealisticcasewherethetrainingdataitisex-posedtoisbiasedinadifferentwaythanthetestdata.1IntroductionManynaturallanguageunderstandingsituationsre-quirereasoningwithrespecttonumbersorquanti-∗MostoftheworkwasdonewhentheauthorswereattheUniversityofIllinois,UrbanaChampaign.ties–understandingfinancialnews,sportsresults,orthenumberofcasualtiesinabombing.Mathwordproblemsformanaturalabstractiontoalotofthesequantitativereasoningproblems.Conse-quently,therehasbeenagrowinginterestindevel-opingautomatedmethodstosolvemathwordprob-lems(Kushmanetal.,2014;Hosseinietal.,2014;RoyandRoth,2015).ArithmeticWordProblemMrs.Hiltbakedpieslastweekendforaholidaydin-ner.Shebaked16pecanpiesand14applepies.Ifshewantstoarrangeallofthepiesinrowsof5pieseach,howmanyrowswillshehave?Solution(16+14)/5=6MathConceptneededforEachOperationFigure1:Anexamplearithmeticwordproblemanditssolution,alongwiththeconceptsrequiredtogenerateeachoperationofthesolutionUnderstandingandsolvingmathwordproblemsinvolvesinterpretingthenaturallanguagedescrip-tionofmathematicalconcepts,aswellasunder-standingtheirinteractionwiththephysicalworld.ConsidertheelementaryschoollevelarithmeticwordproblemshowninFig1.Tosolvetheprob-lem,oneneedstounderstandthat“applepies”and“pecanpies”arekindsof“pies”,andhence,the l D o w n o a d e d f r o m h t t p : / / 直接的

计算语言学协会会刊, 卷. 6, PP. 133–144, 2018. 动作编辑器: Stefan Riezler. 提交批次: 6/2017; 修改批次: 9/2017; 已发表 2/2018. 2018 计算语言学协会. 根据 CC-BY 分发 4.0 执照. C (西德:13) LearningRepresentationsSpecializedinSpatialKnowledge:LeveragingLanguageandVisionGuillemCollellDepartmentofComputerScienceKULeuven3001Heverlee,Belgiumgcollell@kuleuven.beMarie-FrancineMoensDepartmentofComputerScienceKULeuven3001Heverlee,Belgiumsien.moens@cs.kuleuven.beAbstractSpatialunderstandingiscrucialinmanyreal-worldproblems,yetlittleprogresshasbeenmadetowardsbuildingrepresentationsthatcapturespatialknowledge.Here,wemoveonestepforwardinthisdirectionandlearnsuchrepresentationsbyleveragingataskconsistinginpredictingcontinuous2Dspa-tialarrangementsofobjectsgivenobject-relationship-objectinstances(e.g.,“catunderchair”)andasimpleneuralnetworkmodelthatlearnsthetaskfromannotatedimages.Weshowthatthemodelsucceedsinthistaskand,furthermore,thatitiscapableofpredictingcorrectspatialarrangementsforunseenob-jectsifeitherCNNfeaturesorwordembed-dingsoftheobjectsareprovided.Thediffer-encesbetweenvisualandlinguisticfeaturesarediscussed.Next,toevaluatethespatialrepresentationslearnedintheprevioustask,weintroduceataskandadatasetconsistinginasetofcrowdsourcedhumanratingsofspatialsimilarityforobjectpairs.WefindthatbothCNN(convolutionalneuralnetwork)featuresandwordembeddingspredicthumanjudgmentsofsimilaritywellandthatthesevectorscanbefurtherspecializedinspatialknowledgeifweupdatethemwhentrainingthemodelthatpredictsspatialarrangementsofobjects.Overall,thispaperpavesthewaytowardsbuildingdistributedspatialrepresen-tations,contributingtotheunderstandingofspatialexpressionsinlanguage.1IntroductionRepresentingspatialknowledgeisinstrumentalinanytaskinvolvingtext-to-sceneconversionsuchasrobotunderstandingofnaturallanguagecommands(Guadarramaetal.,2013;MoratzandTenbrink,2006)oranumberofrobotnavigationtasks.Despiterecentadvancesinbuildingspecializedrepresenta-tionsindomainssuchassentimentanalysis(Tangetal.,2014),semanticsimilarity/relatedness(Kielaetal.,2015)ordependencyparsing(Bansaletal.,2014),littleprogresshasbeenmadetowardsbuild-ingdistributedrepresentations(a.k.a.embeddings)specializedinspatialknowledge.Intuitively,onemayreasonablyexpectthatthemoreattributestwoobjectsshare(e.g.,size,func-tionality,ETC。),themorelikelytheyaretoexhibitsimilarspatialarrangementswithrespecttootherobjects.Leveragingthisintuition,weforeseethatvisualandlinguisticrepresentationscanbespatiallyinformativeaboutunseenobjectsastheyencodefeatures/attributesofobjects(CollellandMoens,2016).Forinstance,withouthavingeverseenan“elephant”before,butonlya“horse”,onewouldprobablydevisethe“elephant”carryingthe“hu-man”thanotherwise,justbyconsideringtheirsizeattribute.Similarly,onecaninferthata“tablet”anda“book”willshowsimilarspatialpatterns(usuallyonatable,insomeone’shands,ETC。)althoughtheybarelyshowanyvisualresemblance—yettheyaresimilarinsizeandfunctionality.Inthispaperwesystematicallystudyhowinformativevisualandlin-guisticfeatures—intheformofconvolutionalneuralnetwork(CNN)featuresandwordembeddings—areaboutthespatialbehaviorofobjects.Animportantgoalofthisworkistolearndis-tributedrepresentationsspecializedinspatialknowl-edge.Asavehicletolearnspatialrepresentations, l D o w n o a d e d f r o m h t t p : / / 直接的

计算语言学协会会刊, 卷. 6, PP. 121–132, 2018. 动作编辑器: Ani Nenkova. 提交批次: 11/2016; 修改批次: 3/2017; 已发表 2/2018. 2018 计算语言学协会. 根据 CC-BY 分发 4.0 执照. C (西德:13) ConversationModelingonRedditUsingaGraph-StructuredLSTMVictoriaZayatsElectricalEngineeringDepartmentUniversityofWashingtonvzayats@uw.eduMariOstendorfElectricalEngineeringDepartmentUniversityofWashingtonostendor@uw.eduAbstractThispaperpresentsanovelapproachformod-elingthreadeddiscussionsonsocialmediausingagraph-structuredbidirectionalLSTM(long-shorttermmemory)whichrepresentsbothhierarchicalandtemporalconversationstructure.InexperimentswithataskofpredictingpopularityofcommentsinRedditdiscussions,theproposedmodeloutperformsanode-independentarchitecturefordifferentsetsofinputfeatures.Analysesshowabene-fittothemodeloverthefullcourseofthedis-cussion,improvingdetectioninbothearlyandlatestages.Further,theuseoflanguagecueswiththebidirectionaltreestateupdateshelpswithidentifyingcontroversialcomments.1IntroductionSocialmediaprovidesaconvenientandwidelyusedplatformfordiscussionsamongusers.Whenthecomment-responselinksarepreserved,thosecon-versationscanberepresentedinatreestructurewherecommentsrepresentnodes,therootistheoriginalpost,andeachnewreplytoapreviouscom-mentisaddedasachildofthatcomment.Someexamplesofpopularserviceswithtree-likestruc-turesincludeFacebook,红迪网,Quora,andStack-Exchange.Figure1showsanexampleconversa-tiononReddit,wherebiggernodesindicatehigherupvotingofacomment.1InserviceslikeTwitter,1Thetoolhttps://whichlight.github.io/reddit-network-viswasusedtoobtainthisvisualiza-tion.Figure1:VisualizationofasamplethreadonReddit.tweetsandtheirretweetscanalsobeviewedasform-ingatreestructure.Whentimestampsareavail-ablewithacontribution,thenodesofthetreecanbeorderedandannotatedwiththatinformation.Thetreestructureisusefulforseeinghowadiscussionunfoldsintodifferentsubtopicsandshowingdiffer-encesinthelevelofactivityindifferentbranchesofthediscussion.Predictingpopularityofcommentsinsocialme-diaisataskofgrowinginterest.Popularityhasbeendefinedintermsofthevolumeofthere-sponse,butwhenthesocialmediaplatformhasamechanismforreaderstolikeordislikecom-ments(或者,upvote/downvote),thenthedifferenceinpositive/negativevotesprovidesamoreinformativescoreforpopularityprediction.Thisdefinitionof l D o w n o a d e d f r o m h t t p : / / 直接的

计算语言学协会会刊, 卷. 6, PP. 107–119, 2018. 动作编辑器: Ivan Titov. 提交批次: 6/2017; 修改批次: 9/2017; 已发表 2/2018. C(西德:13)2018 计算语言学协会. 根据 CC-BY 分发 4.0 执照. EvaluatingtheStabilityofEmbedding-basedWordSimilaritiesMariaAntoniakCornellUniversitymaa343@cornell.eduDavidMimnoCornellUniversitymimno@cornell.eduAbstractWordembeddingsareincreasinglybeingusedasatooltostudywordassociationsinspecificcorpora.However,itisunclearwhethersuchembeddingsreflectenduringpropertiesoflan-guageoriftheyaresensitivetoinconsequentialvariationsinthesourcedocuments.Wefindthatnearest-neighbordistancesarehighlysen-sitivetosmallchangesinthetrainingcorpusforavarietyofalgorithms.Forallmethods,includingspecificdocumentsinthetrainingsetcanresultinsubstantialvariations.Weshowthattheseeffectsaremoreprominentforsmallertrainingcorpora.Werecommendthatusersneverrelyonsingleembeddingmodelsfordistancecalculations,butratheraverageovermultiplebootstrapsamples,especiallyforsmallcorpora.1IntroductionWordembeddingsareapopulartechniqueinnaturallanguageprocessing(自然语言处理)inwhichthewordsinavocabularyaremappedtolow-dimensionalvectors.Embeddingmodelsareeasilytrained—severalimple-mentationsarepubliclyavailable—andrelationshipsbetweentheembeddingvectors,oftenmeasuredviacosinesimilarity,canbeusedtoreveallatentseman-ticrelationshipsbetweenpairsofwords.Wordem-beddingsareincreasinglybeingusedbyresearchersinunexpectedwaysandhavebecomepopularinfieldssuchasdigitalhumanitiesandcomputationalsocialscience(Hamiltonetal.,2016;Heuser,2016;Phillipsetal.,2017).Embedding-basedanalysesofsemanticsimilaritycanbearobustandvaluabletool,butwefindthatstandardmethodsdramaticallyunder-representthevariabilityofthesemeasurements.Embeddingalgo-rithmsaremuchmoresensitivethantheyappeartofactorssuchasthepresenceofspecificdocuments,thesizeofthedocuments,thesizeofthecorpus,andevenseedsforrandomnumbergenerators.Ifusersdonotaccountforthisvariability,theirconclusionsarelikelytobeinvalid.Fortunately,wealsofindthatsimplyaveragingovermultiplebootstrapsamplesissufficienttoproducestable,reliableresultsinallcasestested.NLPresearchinwordembeddingshassofarfo-cusedonadownstream-centeredusecase,wheretheendgoalisnottheembeddingsthemselvesbutperformanceonamorecomplicatedtask.Forexam-ple,wordembeddingsareoftenusedasthebottomlayerinneuralnetworkarchitecturesforNLP(Ben-gioetal.,2003;Goldberg,2017).Theembeddings’trainingcorpus,whichisselectedtobeaslargeaspossible,isonlyofinterestinsofarasitgeneralizestothedownstreamtrainingcorpus.Incontrast,otherresearcherstakeacorpus-centeredapproachanduserelationshipsbetweenem-beddingsasdirectevidenceaboutthelanguageandcultureoftheauthorsofatrainingcorpus(Bolukbasietal.,2016;Hamiltonetal.,2016;Heuser,2016).Embeddingsareusedasiftheyweresimulationsofasurveyaskingsubjectstofree-associatewordsfromqueryterms.Unlikethedownstream-centeredapproach,thecorpus-centeredapproachisbasedondirecthumananalysisofnearestneighborstoembed-dingvectors,andthetrainingcorpusisnotsimplyanoff-the-shelfconveniencebutratherthecentralobjectofstudy. l D o w n o a d e d f r o m h t t p : / / 直接的 . 米

计算语言学协会会刊, 卷. 6, PP. 91–106, 2018. 动作编辑器: Alexander Clark. 提交批次: 7/2017; 修改批次: 10/2017; 已发表 2/2018. 2018 计算语言学协会. 根据 CC-BY 分发 4.0 执照. C (西德:13) TowardsEvaluatingNarrativeQualityInStudentWritingSwapnaSomasundaran1,MichaelFlor1,MartinChodorow2HillaryMolloy3BinodGyawali1LauraMcCulla11EducationalTestingService,660RosedaleRoad,普林斯顿大学,NJ08541,USA2HunterCollegeandtheGraduateCenter,CUNY,NewYork,NY10065,USA3EducationalTestingService,90NewMontgomeryStreet,SanFrancisco,CA94105,USA{ssomasundaran,mflor,hmolloy,bgyawali,LMcCulla}@ets.orgmartin.chodorow@hunter.cuny.eduAbstractThisworklaysthefoundationforautomatedassessmentsofnarrativequalityinstudentwriting.Wefirstmanuallyscoreessaysfornarrative-relevanttraitsandsub-traits,andmeasureinter-annotatoragreement.Wethenexplorelinguisticfeaturesthatareindicativeofgoodnarrativewritingandusethemtobuildanautomatedscoringsystem.Experimentsshowthatourfeaturesaremoreeffectiveinscoringspecificaspectsofnarrativequalitythanastate-of-the-artfeatureset.1IntroductionNarrative,whichincludespersonalexperiencesandstories,realorimagined,isamediumofexpressionthatisusedfromtheveryearlystagesofachild’slife.Narrativesarealsoemployedinvariouscapac-itiesinschoolinstructionandassessment.Forex-ample,theCommonCoreStateStandards,aned-ucationalinitiativeintheUnitedStatesthatdetailsrequirementsforstudentknowledgeingradesK-12,employsliterature/narrativesasoneofitsthreelanguageartsgenres.Withtheincreasedfocusonautomatedevaluationofstudentwritingineduca-tionalsettings(Adams,2014),automatedmethodsforevaluatingnarrativeessaysatscalearebecomingincreasinglyimportant.Automatedscoringofnarrativeessaysisachal-lengingarea,andonethathasnotbeenexploredex-tensivelyinNLPresearch.Previousworkonauto-matedessayscoringhasfocusedoninformational,争论的,persuasiveandsource-basedwritingconstructs(StabandGurevych,2017;NguyenandLitman,2016;Farraetal.,2015;Somasundaranetal.,2014;BeigmanKlebanovetal.,2014;ShermisandBurstein,2013).相似地,operationalessayscoringengines(AttaliandBurstein,2006;艾略特,2003)aregearedtowardsevaluatinglanguageprofi-ciencyingeneral.Inthiswork,welaytheground-workandpresentthefirstresultsforautomatedscor-ingofnarrativeessays,focusingonnarrativequality.Oneofthechallengesinnarrativequalityanal-ysisisthescarcityofscoredessaysinthisgenre.Wedescribeadetailedmanualannotationstudyonscoringstudentessaysalongmultipledimensionsofnarrativequality,suchasnarrativedevelopmentandnarrativeorganization.UsingascoringrubricadaptedfromtheU.S.CommonCoreStateStan-dards,weannotated942essayswrittenfor18differ-entessay-promptsbystudentsfromthreedifferentgradelevels.Thisdatasetprovidesavarietyofstorytypesandlanguageproficiencylevels.Wemeasuredinter-annotatoragreementtounderstandreliabilityofscoringstoriesfortraits(e.g.,development)aswellassub-traits(e.g.,plotdevelopmentandtheuseofnarrativetechniques).Anumberoftechniquesforwritinggoodstoriesaretargetedbythescoringrubrics.Weimplementedasystemforautomaticallyscoringdifferenttraitsofnarratives,usinglinguisticfeaturesthatcapturesomeofthosetechniques.Weinvestigatedtheeffec-tivenessofeachfeatureforscoringnarrativetraitsandanalyzedtheresultstoidentifysourcesoferrors.Themaincontributionsofthisworkareasfol-lows:(1)Tothebestofourknowledge,thisisthefirstdetailedannotationstudyonscoringnarra-tiveessaysfordifferentaspectsofnarrativequality. l D o w n o a d e d f r o m h t t p : / / 直接的

计算语言学协会会刊, 卷. 6, PP. 77–89, 2018. 动作编辑器: Patrick Pantel. 提交批次: 6/2017; 修改批次: 10/2017; 已发表 2/2018. 2018 计算语言学协会. 根据 CC-BY 分发 4.0 执照. C (西德:13) EventTimeExtractionwithaDecisionTreeofNeuralClassifiersNilsReimers†,NazaninDehghani‡∗,IrynaGurevych††UbiquitousKnowledgeProcessingLab(英国共产党)andResearchTrainingGroupAIPHESDepartmentofComputerScience,TechnischeUniversit¨atDarmstadt‡SchoolofElectricalandComputerEngineering,UniversityofTehranwww.ukp.tu-darmstadt.deAbstractExtractingtheinformationfromtextwhenaneventhappenedischallenging.Documentsdonotonlyreportoncurrentevents,butalsoonpasteventsaswellasonfutureevents.Often,therelevanttimeinformationforaneventisscatteredacrossthedocument.Inthispaperwepresentanovelmethodtoauto-maticallyanchoreventsintime.Toourknowl-edgeitisthefirstapproachthattakestempo-ralinformationfromthecompletedocumentintoaccount.Wecreatedadecisiontreethatappliesneuralnetworkbasedclassifiersatitsnodes.Weusethistreetoincrementallyinfer,inastepwisemanner,atwhichtimeframeaneventhappened.WeevaluatetheapproachontheTimeBank-EventTimeCorpus(Reimersetal.,2016)achievinganaccuracyof42.0%com-paredtoaninter-annotatoragreement(国际航空协会)of56.7%.Foreventsthatspanoverasingledayweobserveanaccuracyimprovementof33.1pointscomparedtothestate-of-the-artCAEVOsystem(Chambersetal.,2014).Withoutre-training,weapplythismodeltotheSemEval-2015Task4onautomatictimelinegenerationandachieveanimprovementof4.01pointsF1-scorecomparedtothestate-of-the-art.Ourcodeispublicallyavailable.11IntroductionKnowingwhenaneventhappenedisusefulforalotofusecases.Examplesareinthefieldsoftime-awareinformationretrieval,textsummarization,automatedtimelinegeneration,andautomaticknowledgebasepopulation.Manyfactsinaknowledgebaseare∗Duringauthor’sinternshipintheresearchtraininggroupAIPHESatUKPLab,TUDarmstadt.1https://github.com/ukplab/tacl2017-event-time-extractiononlytrueforacertaintimeperiod,forexamplethepresidencyofaperson.Hence,thepopulationofaknowledgebasecanhighlybenefitfromhighqualityeventandeventtime2extraction(Surdeanu,2013).Inherenttoeventsistheconnectiontotime.Allan(2002)definesaneventas“somethingthathappensatsomespecifictimeandplace”.Thechallengesforautomaticeventtimeextractionaremanifold.Thetemporalinformationinnewsarticleswhichstateswhenaneventhappenedis,inmostcases,notinthesameorinneighboringsentenceswiththeevent(Reimersetal.,2016).Itcanbementionedfarbeforetheeventorfaraftertheevent.Evenworse,formorethan60%ofevents,thespecificdayatwhichtheeventhappenedisnotmentioned.However,fromtheworldknowledgeandcausalrelations,thereadercaninferalotoftemporalinformationaboutthoseeventsandcanofteninferthattheeventhappenedbeforeoraftersomespecificpointintime.Inthispaperwedescribeanewclassifierforauto-maticeventtimeextraction.WeusetheTimeBank-EventTimeCorpus(Reimersetal.,2016)totrainandevaluateourproposedarchitecture.Incontrasttoothercorporaontemporalrelations,theannota-tionoftheTimeBank-EventTimeCorpusdoesnotmakerestrictionswhere,andinwhichform,tempo-ralinformationforaneventmustbeprovided.Theannotatorswereallowedtotakethewholedocumentintoaccountandwereaskedtoanswer,tothebestoftheirability,thequestionatwhichdateortimeperiodtheeventhappened.Theeventtimeannotationforsomesampleeventsisshowninthefollowing:•Hewas[发送]1980-05-26intospaceonMay26,2Wewillrefertothetemporalinformationwhenaneventhappenedaseventtime. l D o w n o a d e d f r o m h t t p : / / 直接的

计算语言学协会会刊, 卷. 6, PP. 33–48, 2018. 动作编辑器: Regina Barzilay. 提交批次: 5/2016; 修改批次: 10/2016; 已发表 1/2018. 2018 计算语言学协会. 根据 CC-BY 分发 4.0 执照. C (西德:13) JointSemanticSynthesisandMorphologicalAnalysisoftheDerivedWordRyanCotterellDepartmentofComputerScienceJohnsHopkinsUniversityryan.cotterell@jhu.eduHinrichSch¨utzeCISLMUMunichinquiries@cislmu.orgAbstractMuchlikesentencesarecomposedofwords,wordsthemselvesarecomposedofsmallerunits.Forexample,theEnglishwordquestionablycanbeanalyzedasquestion+able+ly.However,thisstructuraldecompositionoftheworddoesnotdirectlygiveusasemanticrepresentationoftheword’smeaning.Sincemorphologyobeystheprincipleofcompositionality,thesemanticsofthewordcanbesystematicallyderivedfromthemeaningofitsparts.Inthiswork,weproposeanovelprobabilisticmodelofwordformationthatcapturesboththeanalysisofawordwintoitsconstituentsegmentsandthesynthesisofthemeaningofwfromthemean-ingsofthosesegments.Ourmodeljointlylearnstosegmentwordsintomorphemesandcomposedistributionalsemanticvectorsofthosemorphemes.WeexperimentwiththemodelonEnglishCELEXdataandGermanDErivBase(Zelleretal.,2013)data.WeshowthatjointlymodelingsemanticsincreasesbothsegmentationaccuracyandmorphemeF1bybetween3%and5%.Additionally,weinvestigatedifferentmodelsofvectorcompo-sition,showingthatrecurrentneuralnetworksyieldanimprovementoversimpleadditivemodels.Finally,westudythedegreetowhichtherepresentationscorrespondtoalinguist’snotionofmorphologicalproductivity.1IntroductionInmostlanguages,wordsdecomposefurtherintosmallerunits,termedmorphemes.Forexample,theEnglishwordquestionablycanbeanalyzedasquestion+able+ly.Thisstructuraldecompositionoftheword,然而,byitselfisnotasemanticrep-resentationoftheword’smeaning;1wefurtherre-quireanaccountofhowtosynthesizethemeaningfromthedecomposition.Fortunately,words—justlikephrases—toalargeextentobeytheprincipleofcompositionality:thesemanticsofthewordcanbesystematicallyderivedfromthemeaningofitsparts.2Inthiswork,weproposeanoveljointprob-abilisticmodelofwordformationthatcapturesbothstructuraldecompositionofawordwintoitscon-stituentsegmentsandthesynthesisofw’smeaningfromthemeaningofthosesegments.Morphologicalsegmentationisastructuredpre-dictiontaskthatseekstobreakawordupintoitsconstituentmorphemes.Theoutputsegmentationhasbeenshowntoaidadiversesetofapplications,suchasautomaticspeechrecognition(Afifyetal.,2006),keywordspotting(Narasimhanetal.,2014),machinetranslation(CliftonandSarkar,2011)andparsing(SeekerandC¸etino˘glu,2015).Incontrasttomuchofthispriorwork,wefocusonsupervisedsegmentation,i.e.,weprovidethemodelwithgoldsegmentationsduringtrainingtime.Insteadofsur-1Therearemanydifferentlinguisticandcomputationaltheo-riesforinterpretingthestructuraldecompositionofaword.Forexample,un-oftensignifiesnegationanditseffectonsemanticscanthenbemodeledbytheoriesbasedonlogic.Thisworkad-dressesthequestionofstructuraldecompositionandsemanticsynthesisinthegeneralframeworkofdistributionalsemantics.2Morphologicalresearchintheoreticalandcomputationallinguisticsoftenfocusesonnoncompositionalorlesscom-positionalphenomena—simplybecausecompositionalderiva-tionposesfewerinterestingresearchproblems.Itisalsotruethat—justasmanyfrequentmultiwordunitsarenotcompletelycompositional—manyfrequentderivations(e.g.,refusal,fit-ness)arenotcompletelycompositional.Anindicationthatnon-lexicalizedderivationsareusuallycompositionalisthefactthatstandarddictionarieslikeOUPeditors(2010)listderivationalaffixeswiththeircompositionalmeaning,withoutahedgethattheycanalsooccuraspartofonlypartiallycompositionalforms.SeealsoHaspelmathandSims(2013),§5.3.6. l D o w n o a d e d f r o m h t t p : / / 直接的

计算语言学协会会刊, 卷. 6, PP. 17–31, 2018. 动作编辑器: Ani Nenkova. 提交批次: 7/17; 修改批次: 11/2017; 已发表 1/2018. 2018 计算语言学协会. 根据 CC-BY 分发 4.0 执照. C (西德:13) MultipleInstanceLearningNetworksforFine-GrainedSentimentAnalysisStefanosAngelidisandMirellaLapataInstituteforLanguage,CognitionandComputationSchoolofInformatics,UniversityofEdinburgh10CrichtonStreet,EdinburghEH89ABs.angelidis@ed.ac.uk,mlap@inf.ed.ac.ukAbstractWeconsiderthetaskoffine-grainedsenti-mentanalysisfromtheperspectiveofmulti-pleinstancelearning(MIL).Ourneuralmodelistrainedondocumentsentimentlabels,andlearnstopredictthesentimentoftextseg-ments,i.e.sentencesorelementarydiscourseunits(EDUs),withoutsegment-levelsupervi-sion.Weintroduceanattention-basedpolar-ityscoringmethodforidentifyingpositiveandnegativetextsnippetsandanewdatasetwhichwecallSPOT(asshorthandforSegment-levelPOlariTyannotations)forevaluatingMIL-stylesentimentmodelslikeours.Experimen-talresultsdemonstratesuperiorperformanceagainstmultiplebaselines,whereasajudge-mentelicitationstudyshowsthatEDU-levelopinionextractionproducesmoreinformativesummariesthansentence-basedalternatives.1IntroductionSentimentanalysishasbecomeafundamentalareaofresearchinNaturalLanguageProcessingthankstotheproliferationofuser-generatedcontentintheformofonlinereviews,博客,internetforums,andsocialmedia.Aplethoraofmethodshavebeenpro-posedintheliteraturethatattempttodistillsenti-mentinformationfromtext,allowingusersandser-viceproviderstomakeopinion-drivendecisions.Thesuccessofneuralnetworksinavarietyofap-plications(Bahdanauetal.,2015;LeandMikolov,2014;Socheretal.,2013)andtheavailabilityoflargeamountsoflabeleddatahaveledtoanin-creasedfocusonsentimentclassification.Super-visedmodelsaretypicallytrainedondocuments(JohnsonandZhang,2015A;JohnsonandZhang,2015乙;Tangetal.,2015;Yangetal.,2016),sen-tences(Kim,2014),orphrases(Socheretal.,2011;[Rating:??]IhadaverymixedexperienceatTheStand.Theburgerandfriesweregood.Thechocolateshakewasdivine:richandcreamy.Thedrive-thruwashorrible.Ittookusatleast30minutestoorderwhentherewereonlyfourcarsinfrontofus.Wecomplainedaboutthewaitandgotahalf–heartedapology.Iwouldgobackbecausethefoodisgood,butmyonlyhesitationisthewait.Summary+Theburgerandfriesweregood+Thechocolateshakewasdivine+Iwouldgobackbecausethefoodisgood–Thedrive-thruwashorrible–Ittookusatleast30minutestoorderFigure1:AnEDU-basedsummaryofa2-out-of-5starreviewwithpositiveandnegativesnippets.Socheretal.,2013)annotatedwithsentimentla-belsandusedtopredictsentimentinunseentexts.Coarse-graineddocument-levelannotationsarerel-ativelyeasytoobtainduetothewidespreaduseofopiniongradinginterfaces(e.g.,starratingsac-companyingreviews).Incontrast,theacquisitionofsentence-orphrase-levelsentimentlabelsre-mainsalaboriousandexpensiveendeavordespiteitsrelevancetovariousopinionminingapplica-tions,e.g.,detectingorsummarizingconsumeropin-ionsinonlineproductreviews.Theusefulnessoffiner-grainedsentimentanalysisisillustratedintheexampleofFigure1,wheresnippetsofopposingpo-laritiesareextractedfroma2-starrestaurantreview.Although,asawhole,thereviewconveysnegativesentiment,aspectsofthereviewer’sexperiencewereclearlypositive.Thisgoeslargelyunnoticedwhenfocusingsolelyonthereview’soverallrating.Inthiswork,weconsidertheproblemofsegment-levelsentimentanalysisfromtheperspectiveofMultipleInstanceLearning(MIL;Keeler,1991). l D o w n o a d e d f r o m h t t p : / / 直接的