Proceedings of the 2nd Workshop on Computational Approaches to Subjectivity and Sentiment ...

210 Pages·2011·5.36 MB·English

ACL HLT 2011
Workshop on Computational Approaches to
Subjectivity and Sentiment Analysis
WASSA
Proceedings of the Workshop
24 June, 2011
Portland, Oregon, USA
ProductionandManufacturingby
Omnipress,Inc.
2600AndersenStreet
Madison,WI53704USA
EndorsedbySIGANN(ACLSpecialInterestGroupforAnnotation)
EndorsedbySIGNLL(ACL’sSpecialInterestGrouponNaturalLanguageLearning)
Sponsored by the Academic Institute for Research in Computer Science (Instituto Universitario de
Investigacio´nInforma´tica),UniversityofAlicante,Spain
Partially funded by the Spanish Ministry of Science and Education of the Spanish Government
(MinisteriodeCienciaeInnovacio´n-GobiernodeEspan˜a)throughtheTIN2009-13391-C04-01grant
Partially funded by the Education Council of the Valencian Community (Conselleria d’Educacio´ -
GeneralitatValenciana),throughthePROMETEO/2009/119andACOMP/2010/286grants
(cid:13)c 2011TheAssociationforComputationalLinguistics
OrdercopiesofthisandotherACLproceedingsfrom:
AssociationforComputationalLinguistics(ACL)
209N.EighthStreet
Stroudsburg,PA18360
USA
Tel: [phone]
Fax: [phone]
[email protected]
ISBN-139781937284060
ii
Foreword
Recent years have marked the beginning and expansion of the Social Web, in which people freely
express and respond to opinion on a whole variety of topics. While the growing volume of subjective
informationavailableallowsforbetterandmoreinformeddecisionsoftheusers,thequantityofdatato
beanalyzedimposedthedevelopmentofspecializedNaturalLanguageProcessing(NLP)systemsthat
automaticallydetectsubjectivityandsentimentintextandsubsequentlyextract,classifyandsummarize
the opinions available on different topics. Although the subjectivity and sentiment analysis research
fields have been highly dynamic in the past years, dealing with subjectivity and sentiment in text has
proven to be a complex, interdisciplinary problem that remains far from being solved. Its challenges
include the need to address the issue from different perspectives and at different levels, depending on
the characteristics of the textual genre, the language(s) treated and the final application for which the
analysisisdone.
Inspired by the objectives we aimed at in the first edition of the Workshop on Computational
Approaches to Subjectivity Analysis (WASSA 2010) and the final outcome, the purpose of the
secondeditionoftheWorkshoponComputationalApproachestoSubjectivityandSentimentAnalysis
(WASSA 2.011) was to create a framework for presenting and discussing the challenges related
to subjectivity and sentiment analysis in NLP, from an interdisciplinary theoretical and practical
perspective.
WASSA 2.011 was organized in conjunction to the 49th Annual Meeting of the Association for
Computational Linguistics: Human Language Technologies, on June 24, 2011, in Portland, Oregon,
U.S.A.Wereceivedatotalof51submissions,fromawiderangeofcountries,ofwhich9wereaccepted
asfullpapers(17%)andanother15asshortpapers(29%). Eachpaperhasbeenreviewedby2members
oftheProgramCommittee. Theacceptedpaperswereallhighlyassessedbythereviewingcommittee,
thebestpaperreceivinganaveragepunctuationof4.5outof5.
The main topics of the accepted papers are the creation, annotation and evaluation of resources for
subjectivityandsentimentanalysisinamonolingual,cross-lingualandmultilingualsetting,subjectivity
andsentimentanalysisindifferenttexttypesandatdifferentlevelsofgranularity. Additionally,WASSA
2.011authorshavecontemplatedinterdisciplinaryanalyses,concerningthegender-specificityanalysis
insubjectivetexts,therelationbetweensentimentandsubjectivityanalysiswithsocialnetworkmining,
opinionquestionansweringandemotiondetection.
The invited talks reflected the interdisciplinary nature of the research in affect-related phenomena as
well. Prof. Jonathan Gratch, from the Universityof Southern California presented a talkon “Emotion
theories,modelsandtheirrelevancetosentimentanalysis”,fromamoregeneralArtificialIntelligence
perspective. Prof. ClaireCardiegaveatalkonthechallengesrelatedtotheimplementationofsentiment
analysissystemsinreal-worldapplications.
Giventhedemonstratedandincreasinglygrowinginterestinthetopicsaddressed,wehopethatWASSA
will continue to be organized in the next years and become an established forum for researchers to
discussanddebatethebestpracticesinsubjectivityandsentimentanalysis.
WewouldliketothanktheACL-HLT2011Organizersforthehelpandsupportatthedifferentstagesof
iii
theworkshoporganizationprocess. WearealsoespeciallygratefultotheProgramCommitteemembers
and the external reviewers for the time and effort spent assessing the papers. We would like to extend
our thanks to our invited speakers – Prof. Jonathan Gratch and Prof. Claire Cardie, for accepting to
deliverthekeynotetalks.
Secondly, we would like to express our gratitude for the official endorsement we received from
SIGANN (ACL Special Interest Group for Annotation) and SIGNLL (ACL Special Interest Group
onNaturalLanguageLearning).
Further on, we would like tothank the Editors of the Decision SupportSystems Journal, published by
Elsevier, for accepting to organize a Special Issue of this journal containing the extended versions of
theWASSA2.011fullpapers.
We would like to express our gratitude to the team at the Department of Software and Computing
SystemsattheUniversityofAlicante-JavierFerna´ndez,whocreatedtheWASSAlogoandtoMiguel
A´ngelVaroandMiguelA´ngelBaeza-forthetechnicalsupporttheyprovided.
Last, but not least, we are grateful for the financial support given by Academic Institute for Research
in Computer Science of the University of Alicante (Instituto Universitario para la Investigacio´n en
Informa´tica, Universidad de Alicante), the Spanish Ministry of Science and Education of the Spanish
Government(MinisteriodeCienciaeInnovacio´n-GobiernodeEspan˜a)throughtheTIN2009-13391-
C04-01 grant, and to the Education Council of the Valencian Community (Conselleria d’Educacio´ -
GeneralitatValenciana),throughthePROMETEO/2009/119andACOMP/2010/286grants.
AlexandraBalahur,EsterBoldrini,Andre´sMontoyo,PatricioMart´ınez-Barco
WASSA2.011Chairs
iv
Organizers:
AlexandraBalahur-UniversityofAlicante,Spain
EsterBoldrini-UniversityofAlicante,Spain
Andre´sMontoyo-UniversityofAlicante,Spain
PatricioMart´ınez-Barco-UniversityofAlicante,Spain
ProgramCommittee:
EnekoAgirre-UniversityoftheBasqueCountry,Spain
NicolettaCalzolari-CNRPisa,Italy
ErikCambria-UniversityofStirling,U.K.
Jose´ CarlosCortizo-EuropeanUniversityMadrid,Spain
Jesu´sM.Hermida-UniversityofAlicante,Spain
VeroniqueHoste-UniversityofGhent,Belgium
MijailKabadjov-EC-JointResearchCentre,Italy
ZornitsaKozareva-InformationSciencesInstitute,U.S.A.
RadaMihalcea-UniversityofNorthTexas,U.S.A.
RafaelMun˜oz-UniversityofAlicante,Spain
Gu¨nterNeumann-DFKI,Germany
ConstantinOrasan-UniversityofWolverhampton,U.K.
ManuelPalomar-UniversityofAlicante,Spain
ViktorPekar-UniversityofWolverhampton,U.K.
PaoloRosso-PolytechnicUniversityofValencia,Spain
JosefSteinberger-EC-JointResearchCentre,Italy
RalfSteinberger-EC-JointResearchCentre,Italy
VeselynStoyanov-CornellUniversity,U.S.A.
CarloStrapparava-FBK,Italy
MaiteTaboada-SimonFraserUniversity,Canada
HristoTanev-EC-JointResearchCentre,Italy
MikeThelwall-UniversityofWolverhampton,U.K.
Jose´ AntonioTroyano-UniversityofSeville,Spain
DanTufis-RACAI,Romania
AlfonsoUren˜a-UniversityofJae´n,Spain
TarasZagibalov-Brandwatch,U.K.
AdditionalReviewers:
ElenaLloret-UniversityofAlicante,Spain
Mar´ıa-TeresaMart´ınValdivia-UniversityofJae´n,Spain
SaifMohammad-NationalResearchCouncil,Canada
v
InvitedSpeakers:
ClaireCardie-CornellUniversity,U.S.A.-Appinions,U.S.A.
JonathanGratch-UniversityofSouthernCalifornia,U.S.A.
vi
Table of Contents
CatsRuleandDogsDrool!: ClassifyingStanceinOnlineDebate
Pranav Anand, Marilyn Walker, Rob Abbott, Jean E. Fox Tree, Robeson Bowmani and Michael
Minor......................................................................................1
Averblexiconmodelfordeepsentimentanalysisandopinionminingapplications
IsaMaksandPiekVossen..............................................................10
ExperimentswithaDifferentialSemanticsAnnotationforWordNet3.0
DanTufisandDanStefanescu..........................................................19
CreatingSentimentDictionariesviaTriangulation
Josef Steinberger, Polina Lenkova, Mohamed Ebrahim, Maud Ehrman, Ali Hurriyetoglu, Mijail
Kabadjov,RalfSteinberger,HristoTanev,VanniZavarellaandSilviaVazquez....................28
GeneratingSemanticOrientationLexiconusingLargeDataandThesaurus
AmitGoyalandHalDaume............................................................37
DevelopingRobustModelsforFavourabilityAnalysis
DaoudClarke,PeterLaneandPaulHender...............................................44
DetectingImplicitExpressionsofSentimentinTextBasedonCommonsenseKnowledge
AlexandraBalahur,Jesu´sM.HermidaandAndre´sMontoyo...............................53
ALinktothePast: ConstructingHistoricalSocialNetworks
MatjevandeCampandAntalvandenBosch.............................................61
TrackingSentimentinMail: HowGendersDifferonEmotionalAxes
SaifMohammadandTonyYang ........................................................ 70
DevelopingJapaneseWordNetAffectforAnalyzingEmotions
YoshimitsuTorii,DipankarDas,SivajiBandyopadhyayandManabuOkumura .............. 80
ImprovingaMethodforQuantifyingReaders’ImpressionsofNewsArticleswithaRegressionEquation
TadahikoKumamoto,YukikoKawaiandKatsumiTanaka ................................. 87
FeatureSelectionforSentimentAnalysisBasedonContentandSyntaxModels
AdnanDuricandFeiSong ............................................................. 96
AutomaticEmotionClassificationforInterpersonalCommunication
FrederikVaassenandWalterDaelemans................................................104
AutomaticSentimentClassificationofProductReviewsUsingMaximalPhrasesBasedAnalysis
MariaTchalakova,DaleGerdemannandDetmarMeurers................................111
MiningSubjectiveKnowledgefromCustomerReviews: ASpecificCaseofIronyDetection
AntonioReyesandPaoloRosso ....................................................... 118
vii
AutomaticExpansionofFeature-LevelOpinionLexicons
Ferm´ınL.Cruz,Jose´ A.Troyano,F.JavierOrtegaandFernandoEnr´ıquez ................. 125
RobustSense-basedSentimentClassification
BalamuraliAR,AdityaJoshiandPushpakBhattacharyya ................................ 132
SentimentClassificationUsingSemanticFeaturesExtractedfromWordNet-basedResources
YoanGutie´rrez,SoniaVa´zquezandAndre´sMontoyo....................................139
OntheDifficultyofClusteringMicroblogTextsforOnlineReputationManagement
FernandoPerez-Tellez,DavidPinto,JohnCardiffandPaoloRosso........................146
EMOCause: AnEasy-adaptableApproachtoExtractEmotionCauseContexts
IreneRusso,TommasoCaselli,FrancescoRubino,EsterBoldriniandPatricioMart´ınez-Barco153
ACross-corpusStudyofUnsupervisedSubjectivityIdentificationbasedonCalibratedEM
DongWangandYangLiu.............................................................161
TowardsaUnifiedApproachforOpinionQuestionAnsweringandSummarization
ElenaLloret,AlexandraBalahur,ManuelPalomarandAndre´sMontoyo...................168
CorporateNewsClassificationandValencePrediction: ASupervisedApproach
SyedAqueelHaiderandRishabhMehrotra ............................................. 175
InstanceLevelTransferLearningforCrossLingualOpinionAnalysis
RuifengXu,JunXuandXiaolongWang................................................182
Sentimatrix–MultilingualSentimentAnalysisService
Alexandru-LucianGinsca,EmanuelaBoros,AdrianIftene,DianaTrandabat,MihaiToader,Mar-
iusCorici,Cenel-AugustoPerezandDanCristea.............................................189
viii
Workshop Program
FridayJune24,2011
(8:45)OpeningRemarks
(9:00)Invitedtalk(I):Prof. JonathanGratch
(9:40)Invitedtalk(II):Prof. ClaireCardie
(10:15)BestPaperAward
CatsRuleandDogsDrool!: ClassifyingStanceinOnlineDebate
PranavAnand, MarilynWalker, RobAbbott, JeanE.FoxTree, RobesonBowmani
andMichaelMinor
(10:40)Break
(11:00)Session1: ResourcesforSentimentAnalysis
Averblexiconmodelfordeepsentimentanalysisandopinionminingapplications
IsaMaksandPiekVossen
ExperimentswithaDifferentialSemanticsAnnotationforWordNet3.0
DanTufisandDanStefanescu
CreatingSentimentDictionariesviaTriangulation
Josef Steinberger, Polina Lenkova, Mohamed Ebrahim, Maud Ehrman, Ali Hur-
riyetoglu,MijailKabadjov,RalfSteinberger,HristoTanev,VanniZavarellaandSil-
viaVazquez
GeneratingSemanticOrientationLexiconusingLargeDataandThesaurus
AmitGoyalandHalDaume
ix
FridayJune24,2011(continued)
(12:30)LunchBreak
(13:30)Session2: ResourcesandApplicationsofSentimentAnalysis
DevelopingRobustModelsforFavourabilityAnalysis
DaoudClarke,PeterLaneandPaulHender
DetectingImplicitExpressionsofSentimentinTextBasedonCommonsenseKnowledge
AlexandraBalahur,Jesu´sM.HermidaandAndre´sMontoyo
ALinktothePast: ConstructingHistoricalSocialNetworks
MatjevandeCampandAntalvandenBosch
TrackingSentimentinMail: HowGendersDifferonEmotionalAxes
SaifMohammadandTonyYang
DevelopingJapaneseWordNetAffectforAnalyzingEmotions
YoshimitsuTorii,DipankarDas,SivajiBandyopadhyayandManabuOkumura
(15:30)Break
(16:00)Session3: SentimentClassification
ImprovingaMethodforQuantifyingReaders’ImpressionsofNewsArticleswithaRegres-
sionEquation
TadahikoKumamoto,YukikoKawaiandKatsumiTanaka
FeatureSelectionforSentimentAnalysisBasedonContentandSyntaxModels
AdnanDuricandFeiSong
AutomaticEmotionClassificationforInterpersonalCommunication
FrederikVaassenandWalterDaelemans
Automatic Sentiment Classification of Product Reviews Using Maximal Phrases Based
Analysis
MariaTchalakova,DaleGerdemannandDetmarMeurers
MiningSubjectiveKnowledgefromCustomerReviews: ASpecificCaseofIronyDetection
AntonioReyesandPaoloRosso
x