社交媒體文本情感分析

1、聲明本學(xué)位論文是我在導(dǎo)師的指導(dǎo)下取得的研究成果，盡我所知，在本學(xué)位論文中，除了加以標(biāo)注和致謝的部分外，不包含其他人己經(jīng)發(fā)表或公布過的研究成果，也不包含我為獲得任何教育機(jī)構(gòu)的學(xué)位或?qū)W歷而使用過的材料。與我一同工作的同事對本學(xué)位論文做出的貢獻(xiàn)均已在論文中作了明確的說明。研究生簽名：婁盎2曰7年；月fj日學(xué)位論文使用授權(quán)聲明南京理工大學(xué)有權(quán)保存本學(xué)位論文的電子和紙質(zhì)文檔，可以借閱或上網(wǎng)公布本學(xué)位論文的部分或全部內(nèi)容，可以向有關(guān)部門或機(jī)構(gòu)送交并

2、授權(quán)其保存、借閱或上網(wǎng)公布本學(xué)位論文的部分或全部內(nèi)容。對于保密論文，按保密的有關(guān)規(guī)定和程序處理。研究生簽名：差i生z。曙年；月f多日萬方數(shù)據(jù)Abstract碩士學(xué)位論文～一———————————————————————————————h————————————————————————————————————————————————————————————————————————————————————一AbstractInrecenty

3、ears，IntemettechnologyisstilldevelopingrapidlyresultinginalargenumberofInternetapplications，includingsocialnetworkingqhesocialnetworkingapplicationshavealargenumberofusersatalltimesinvolvedinthefigures，products，eventsand

4、otherrelatedsocialmediadataThesentimentanalysistechniqueisusedtominethesubjectivesentimentinformationinthetextSentimentanalysisOilmicro—blogastherepresentativeofthesocialmediacalldevelopthepotentialcommercialandsocialval

5、ueandit’Swidelyusedintheapplicationssuchasproductinformationfeedback，commodityrecommendationalgorithm，publicopinionmonitoring，hoteventtrackingandSOonThispapermainlyfocusesontheproblemofsentimentclassificationforsocialmed

6、iaThefirsttwochaptersintroducethecurrentsituationandbasictechnologyofthisproblemindetail’fhen，accordingtotheshortcomingsoftherelatedresearch，weproposeoursentimentclassificationmethodsinthenextthreechapters(1)Weproposease

7、ntimentclassificationmethodbasedonmachinelearningandsemanticrulesfusionWeusesomesemanticrulestOimprovetheprecisionofsentimentorientationmeasurementoftraditionallexicon—basedapproachThenweaddtherulefeaturestothebasicfeatu

8、resetafterconvertingandexpandingthemOurmethodissuperiortothegeneralfusionmethodinbothsentimentanalysisgranularityandfeaturerepresentationExperimentsshowthatthemethodhasachievedagreaterperformanceimprovementWerankedfirsti

9、ntaskofmicro—blogsentimentclassificationinChineseOpinionAnalysisEvaluation2015(COAE20151(2)Thispaperusesthedistantannotatingapproachtohelpsolvethesentimentclassificationproblem，InChapter4，basedonthesentimentlexiconconstr

10、uctionmethodofneuralnetworkmodel，weimprovedthismodelbyusingsemanticrulesandsampleweightsThedictionaryconstructedbytheimprovedmethodoutperformsdictionariesconstructedbyothermethodsOrmanualannotationWewonthefirstplaceintas

11、kofsentimentwordsextractioninChineseOpinionAnalysisEvaluation2016(COAE2016)basedonthisdictionary(3)WeproposetoimprovetheperformanceofclassificationbyensemblelearningmethodswithdistantsupervisedsocialmediadataFirst，theBag

12、gingensemblelearningmodelisprovedtobesuperiortoanysinglemodelinstabilityandgeneralizationabilityOnthisbasis，theStackingensemblelearningmodelisproposedThismodelcombinesthedistantsuperviseddatawiththemanuallysuperviseddata