Table Of Content

Studies in Big Data 113 Tzung-Pei Hong Leticia Serrano-Estrada Akrati Saxena Anupam Biswas Editors Deep Learning for Social Media Data Analytics Studies in Big Data Volume 113 SeriesEditor JanuszKacprzyk,PolishAcademyofSciences,Warsaw,Poland Theseries“StudiesinBigData”(SBD)publishesnewdevelopmentsandadvances in the various areas of Big Data-quickly and with a high quality. The intent is to coverthetheory,research,development,andapplicationsofBigData,asembedded inthefieldsofengineering,computerscience,physics,economicsandlifesciences. The books of the series refer to the analysis and understanding of large, complex, and/or distributed data sets generated from recent digital sources coming from sensorsorotherphysicalinstrumentsaswellassimulations,crowdsourcing,social networks or other internet transactions, such as emails or video click streams and other. The series contains monographs, lecture notes and edited volumes in Big Data spanning the areas of computational intelligence including neural networks, evolutionary computation, soft computing, fuzzy systems, as well as artificial intelligence, data mining, modern statistics and Operations research, as well as self-organizing systems. Of particular value to both the contributors and the readership are the short publication timeframe and the world-wide distribution, whichenablebothwideandrapiddisseminationofresearchoutput. Thebooksofthisseriesarereviewedinasingleblindpeerreviewprocess. IndexedbySCOPUS,EICompendex,SCIMAGOandzbMATH. AllbookspublishedintheseriesaresubmittedforconsiderationinWebofScience. · · Tzung-Pei Hong Leticia Serrano-Estrada · Akrati Saxena Anupam Biswas Editors Deep Learning for Social Media Data Analytics Editors Tzung-PeiHong LeticiaSerrano-Estrada DepartmentofComputerScience UrbanDesignandRegionalPlanningUnit andInformationEngineering UniversityofAlicante NationalUniversityofKaohsiung Alicante,Spain Kaohsiung,Taiwan AnupamBiswas AkratiSaxena DepartmentofComputerScience DepartmentofMathematicsandComputer andEngineering Science NationalInstituteofTechnologySilchar EindhovenUniversityofTechnology Cachar,Assam,India Eindhoven,TheNetherlands ISSN 2197-6503 ISSN 2197-6511 (electronic) StudiesinBigData ISBN 978-3-031-10868-6 ISBN 978-3-031-10869-3 (eBook) https://doi.org/10.1007/978-3-031-10869-3 ©TheEditor(s)(ifapplicable)andTheAuthor(s),underexclusivelicensetoSpringerNature SwitzerlandAG2022 Thisworkissubjecttocopyright.AllrightsaresolelyandexclusivelylicensedbythePublisher,whether thewholeorpartofthematerialisconcerned,specificallytherightsoftranslation,reprinting,reuse ofillustrations,recitation,broadcasting,reproductiononmicrofilmsorinanyotherphysicalway,and transmissionorinformationstorageandretrieval,electronicadaptation,computersoftware,orbysimilar ordissimilarmethodologynowknownorhereafterdeveloped. Theuseofgeneraldescriptivenames,registerednames,trademarks,servicemarks,etc.inthispublication doesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevant protectivelawsandregulationsandthereforefreeforgeneraluse. Thepublisher,theauthors,andtheeditorsaresafetoassumethattheadviceandinformationinthisbook arebelievedtobetrueandaccurateatthedateofpublication.Neitherthepublishernortheauthorsor theeditorsgiveawarranty,expressedorimplied,withrespecttothematerialcontainedhereinorforany errorsoromissionsthatmayhavebeenmade.Thepublisherremainsneutralwithregardtojurisdictional claimsinpublishedmapsandinstitutionalaffiliations. ThisSpringerimprintispublishedbytheregisteredcompanySpringerNatureSwitzerlandAG Theregisteredcompanyaddressis:Gewerbestrasse11,6330Cham,Switzerland Preface ThecontentofthisvolumefocusesonboththeoryandpracticalapplicationofDeep Learning for Social Media Data Analytics. Nowadays, social media platforms are overwhelmedbycontentconstantlybeinggeneratedbybillionsofusers.Thesehuge volumesofdataaremostlyunstructuredandthus,whileofferinganenormouspoten- tialtoinfluencebusiness,politics,security,andothersocialaspects,technicalchal- lengesemergethatstillneedtobefullyaddressedbyresearch.Socialmediaposts, sharedintheformoftexts,oftenaccompaniedbyimagesorvideos,haveaheteroge- neousnatureandcomplexstructureduetothewiderangeofembeddedexpressions andfeatures,makingdeeplearningandmodeltrainingratherusefulforprocessing, classifying, analyzing, managing, and storing the large volumes of data involved. This information then becomes resourceful for use on many day-to-day applica- tions. For instance, natural language processing of the text shared in social media allowsdrawingconclusionsoftennecessaryforsentimentanalysisoropinionmining, whereasimagerecognitionmodelsarehelpfulformakingsenseofpostscontaining photoswithnocaption.Likewise,acquiringusers’personalinterestsandidentifying socialmediainfluencersiscrucialfortargetedadvertisements,whereasgroup-centric studies, flow of information, and prediction of links are relevant for recommenda- tions. Effective customer services based on chatbots allow businesses to automate customer services. For instance, interactive question-answering systems are often usedforwhichdeeplearningtechniquesareincorporated. In view of its great potential, the research presented in this book emerges from theconvictionthatthereisstillmuchprogresstobemadetowardsexploitingdeep learninginthecontextofsocialmediadataanalyticsanditsapplicationstobenefit societyinawiderangeofareasfromhealthtosmartcities.Thecollaborativeeffort between experts fromdifferentfields hasenabled theintegration ofavaried range ofapproachestodeeplearning-basedtheories,techniques,andmethodsappliedto socialmediadata.Itcomprisesdetailedstudiesthatcoversomeofthemostsalient areasofresearchinthefield. The book is organized into four Parts, which include 15 chapters that report on original research in network structure analysis, social media text analysis, user behavioranalysis,andsocialmediasecurityanalysis.InPartIofthebook,Chapter v vi Preface “NodeClassificationUsingDeepLearninginSocialNetworks”presentsacompre- hensiveliteraturereviewofdeeplearningresearchmethodsdesignedfornodeclassi- fication,offeringafutureresearchagendainthefield,whereasChapter“NN-LP-CF: NeuralNetworkBasedLinkPredictiononSocialNetworksUsingCentrality-Based Features”proposes aneuralnetwork-based linkpredictionmodelusingcentrality- basedfeaturesandglobaltopologicalfeaturestopredictmissinglinks.Real-world networks are used for performing the experiments, which are then compared to state-of-the-artmethods. PartIIdelvesintotextanalyticstechniquesandapplications.Specifically,Chapter “Deep Learning for Code-Mixed Text Mining in Social Media: A Brief Review” presents a thorough review of deep learning techniques for code-mixed text analysis, introducing a range of text-mining tasks that use deep learning to model different solutions. Some of the challenges associated and the deep learning techniques used to overcome them are outlined. Chapter “Convolutional and Recur- rentNeuralNetworksforOpinionMiningonDrugReviews”focusesonsentiment analysisandopinionminingappliedtodrugreviewsusingdeepconvolutionaland recurrent neural networks and hybrid models for predicting categories on health forumdata.Followingasimilardirection,Chapter“Text-BasedSentimentAnalysis Using Deep Learning Techniques” proposes a comparative experimental analysis of selected deep learning models for conducting a text-based sentiment analysis, specificallyappliedtoanIMDBmoviereviewsdataset.Chapter“SocialSentiment Analysis Using Features Based Intelligent Learning Techniques” offers a comple- mentary approach by analyzing and classifying text from Twitter,Amazon mobile reviews and IMDB movie reviews into two classes, namely positive and negative sentimentsbyadoptingvariousmachinelearningalgorithms. PartIIIisconcernedwiththetheoreticalbackgroundandexperimentalexamples of how deep learning techniques are applied to recommendation systems and user behavioranalysis.Inthissection,Chapter“Modified-PIPwithDeepNeuralNetwork (DNN)Architecture:ACoherentRecommendationFrameworkforCapturingUser Behaviour”seekstodevelopahigh-performancedeeplearningmodelforcomputing similaritybehaviorfrommovierecommendationsystemsusingaMovieLens-100K dataset.Chapter“ASurveyonGraphNeuralNetworkBasedVideoRecommenda- tion System” highlights the key benefits and convenience of deep learning-based recommendersystemsovertraditionalones.Relevantapplicationsandissuesrelated to data privacy are explained and future directions of the field are outlined. In a different approach, the research presented in Chapter “Characterisation of Mental Health Conditions in Social Media Using Deep Learning Techniques” covers a comprehensive review of previous work developed in relation to the characteriza- tion of social media users’ mental health conditions through deep learning techniques and the future directions in this field of knowledge. In line with this topic, Chapter“PredictingMentalHealthandNutritionalStatusfromSocialMediaProfile Using Deep Learning” focuses on the prediction of mental health and nutritional statusofFacebookusersthroughFacebook’sConfessionsDatasets(MentalHealth Datasetcategory),bydesigningandapplyingadeeplearningmodel.Inadifferent approachtouserbehavior,Chapter“ImpactofArtificialIntelligence-BasedChatbots Preface vii onCustomerEngagementandBusinessGrowth”delvesintohowcustomersupport, service, and business growth, which are strongly linked to the client’s perception oftrust,arebeingimpactedbytheintroductionofartificialintelligence-basedchat- boxes. Several case-study companies and brands are analyzed, and a discussion is presentedonhowtheirefforttoautomatecommunicationwithcustomershasresulted inbusinessgrowth. Thelastpartofthebookcoversthestate-of-the-artanddeeplearningtechniques thathaverecentlybeendevelopedtoaddresssomeofthemultipleformsofsecurity- related challenges in social media platforms. Specifically, Chapter “Do Not ‘Fake It Till you Make It’! Synopsis of Trending Fake News Detection Methodologies UsingDeepLearning”focusesonthedetectionoffakecontentinsocialmediaby developingacomparativestudyofrecentfindings,architectures,currentinsights,and discussionsinthefield.Italsoshedssomelightonupcomingtrendsandfuturedirec- tionsonthedevelopment ofmoreefficientfakenewsdetectors.Chapter“Towards DetectingFakeSpammersGroupsinSocialMedia:AnUnsupervisedDeepLearning Approach”includesathoroughreviewoftheapplicationofmultipledeeplearning andmachinelearning-basedunsupervisedalgorithmstodetectfalsereviewswithfake reviewers. Advantages and disadvantages of some of the most popular algorithms arediscussed,andtheimplementationofthesetoaGooglePlayStoreAppdatasetis usedasareal-timecasestudy.Chapter“ADeepLearningApproachforAnomalous User-IntrusionDetectioninSocialMediaNetworkSystem”delvesintosocialmedia networkintrusiondetectionsystemsandproposesanoveldeeplearningapproachfor detectingandclassifyingthetypesofattacks.Chapter“DeepDiggingofAnomalous TransactionsinFinancialNetworkswithImbalancedData”addressesthechallenges involvedinthedetectionofanomaliesorfraudulentusersandtransactions.Specifi- cally,theresearchcoversdeeplearningapproachestofraudulenttransactiondetection inthefinancialnetworkdomain. Allinall,theeditorswouldliketoexpresstheirdeepappreciationandgratitude to the contributors who have made their research available for this volume, and to theanonymousreviewerswhohaveprovidedaninvaluableserviceinreferringtothe chapters.Wesincerelyhopethatthisworkservesasareferenceforresearchers,as wellasacompilationofinnovativeideasandsolutionsforpractitionersinterestedin recenttheoriesandapplicationsofdeeplearningforsocialmediadataanalytics. Sincerely, Kaohsiung,Taiwan Dr.Tzung-PeiHong Alicante,Spain Dr.LeticiaSerrano-Estrada Eindhoven,TheNetherlands Dr.AkratiSaxena Cachar,India Dr.AnupamBiswas April2022 Contents NetworkStructureAnalysis NodeClassificationUsingDeepLearninginSocialNetworks .......... 3 AiktaArya,PradumnKumarPandey,andAkratiSaxena NN-LP-CF:NeuralNetworkBasedLinkPredictiononSocial NetworksUsingCentrality-BasedFeatures .......................... 27 ShashankShesharSingh, DivyaSrivastva, AjayKumar, andVishalSrivastava SocialMediaTextAnalysis DeepLearning forCode-Mixed Text MininginSocialMedia: ABriefReview .................................................... 45 RrubaaPanchendrarajanandAkratiSaxena Convolutional and Recurrent Neural Networks for Opinion MiningonDrugReviews ........................................... 65 NesmaSettoutiandFatihaYoubi Text-BasedSentimentAnalysisUsingDeepLearningTechniques ...... 81 SiddhiKaduandBhartiJoshi Social Sentiment Analysis Using Features Based Intelligent LearningTechniques ............................................... 101 PrasannavenkatesanTheerthagiri UserBehaviourAnalysis Modified-PIPwithDeepNeuralNetwork(DNN)Architecture: ACoherentRecommendationFrameworkforCapturingUser Behaviour ........................................................ 121 BamBahadurSinha, GurvinderSinghYadav, andSagarBadrishKudkelwar ix x Contents ASurveyonGraphNeuralNetworkBasedVideoRecommendation System ........................................................... 141 ToshiRawkaandMahipalJadeja CharacterisationofMentalHealthConditionsinSocialMedia UsingDeepLearningTechniques ................................... 157 ToshitaSharma,RrubaaPanchendrarajan,andAkratiSaxena Predicting Mental Health and Nutritional Status from Social MediaProfileUsingDeepLearning ................................. 177 ZakirHussainandMalayaDuttaBorah ImpactofArtificialIntelligence-BasedChatbotsonCustomer EngagementandBusinessGrowth .................................. 195 ChitraKrishnan,AdityaGupta,AsthaGupta,andGurinderSingh SocialMediaSecurityAnalysis DoNot‘FakeItTillYouMakeIt’!SynopsisofTrendingFake NewsDetectionMethodologiesUsingDeepLearning .................. 213 RishabhMisraandJigyasaGrover Towards Detecting Fake Spammers Groups in Social Media: AnUnsupervisedDeepLearningApproach .......................... 237 JayeshSoni,NagarajanPrabakar,andHimanshuUpadhyay A Deep Learning Approach for Anomalous User-Intrusion DetectioninSocialMediaNetworkSystem ........................... 255 NidhiLal,ShishupalKumar,andGarimaKaidan DeepDiggingofAnomalousTransactionsinFinancialNetworks withImbalancedData ............................................. 277 VaishaliKansalandPradumnPandey