Skip navigation
Please use this identifier to cite or link to this item: https://repository.esi-sba.dz/jspui/handle/123456789/492
Full metadata record
DC FieldValueLanguage
dc.contributor.authorABDELLATIF, ABdelraoufL
dc.contributor.authorKHEMKHAM, MOhamed-
dc.date.accessioned2023-10-15T10:01:36Z-
dc.date.available2023-10-15T10:01:36Z-
dc.date.issued2023-
dc.identifier.urihttps://repository.esi-sba.dz/jspui/handle/123456789/492-
dc.descriptionEncadrant : Dr Sid Ahmed Benabderrahmane / Co-Encadrant : Pr Sidi Mohammed Benslimaneen_US
dc.description.abstractABSTRACT : In this study, we leverage techniques from machine learning, deep learning, natural language processing and sentiment analysis to investigate the presence and impact of biases in the coverage of the Western news platforms and social media sources. Our focus centers around detecting and analyzing biases, differentiating between intentional and accidental misinformation, and interpreting how these biases may influence public perception, followed by sentiment analysis to study the variation of sentiments over time. This research report delves into the intersection of Machine Learning (ML), Deep Learning (DL), Natural Language Processing (NLP), sentiment analysis, and media bias detection. The study elucidates how these innovative technologies can be leveraged to scrutinize and interpret media content, ensuring unbiased information dissemination..*** RÉSUMÉ : Dans cette étude, nous exploitons des techniques issues de l’apprentissage automatique, de l’apprentissage profond, du traitement du langage naturel et de l’analyse des sentiments pour examiner la présence et l’impact des biais dans la couverture des plateformes d’information occidentales et des sources de médias sociaux. Notre attention se concentre sur la détection et l’analyse des biais, en différenciant les informations erronées intentionnelles et accidentelles, et en interprétant comment ces biais peuvent influencer la perception du public, suivis par une analyse des sentiments pour étudier la variation des sentiments dans le temps. Ce mémoire se plonge dans l’intersection de l’Apprentissage Automatique (AA), de l’Apprentissage Profond (AP), du Traitement du Langage Naturel (TLN), de l’analyse des sentiments et de la détection des biais médiatiques. L’étude éclaire comment ces technologies innovantes peuvent être exploitées pour scruter et interpréter le contenu médiatique, assurant une diffusion de l’information sans biais.en_US
dc.language.isoenen_US
dc.subjectArtificial Intelligenceen_US
dc.subjectMachine Learningen_US
dc.subjectDeep Learningen_US
dc.subjectArtificial Neural Networksen_US
dc.subjectNatural Language Processingen_US
dc.subjectSupport Vector Machinesen_US
dc.subjectMean Square Erroren_US
dc.subjectLeast Absolute Shrinkage And Selection Operatoren_US
dc.subjectL1 normsen_US
dc.subjectL2 normsen_US
dc.subjectContinuous Bag Of Wordsen_US
dc.subjectGlobal Vectors For Word Representationen_US
dc.subjectProportional Hazards Regressionen_US
dc.subjectK-Nearest Neighborsen_US
dc.subjectPrincipal Component Analysisen_US
dc.subjectt-Distributed Stochastic Neighbor Embeddingen_US
dc.subjectLong Short-Term Memoryen_US
dc.subjectGated Recurrent Unitsen_US
dc.subjectRadial Basis Function Networksen_US
dc.subjectGenerative Adversarial Networksen_US
dc.subjectSelf-Organizing Mapsen_US
dc.subjectTerm Frequencyen_US
dc.subjectInverse Document Frequencyen_US
dc.subjectTerm Frequency-Inverse Document Frequencyen_US
dc.subjectAutomatic Speech Recognitionen_US
dc.subjectConditional Random Fieldsen_US
dc.subjectBidirectional Long Short-Term Memoryen_US
dc.subjectStanford Question Answering Dataseten_US
dc.subjectQuestion Answeringen_US
dc.subjectNamed Entity Recognitionen_US
dc.subjectDocument Understanding Conferenceen_US
dc.titleBrainwashing 2.0: Bias and double standard in western social media and news outlets.en_US
dc.typeThesisen_US
Appears in Collections:Master

Files in This Item:
File Description SizeFormat 
Master (1)-1-1.pdf73,35 kBAdobe PDFView/Open
Show simple item record


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.