This thesis presents a qualitative study of fake news disseminated in Poland as a result of the outbreak of the war between Russia and Ukraine in February 2022. It aims to contribute to the public knowledge about the most prominent themes of the Polish fake news and the topics prone to be used as disinformation and as a weapon of war in a conflict setting. Through the use of framing theory, this research explains how the conflict in Ukraine has been framed within the Polish fake news. To understand the tactics of the fake news spreaders, culture links and culture pegs used to attract the Polish audience have been identified. The study was conducted using content analysis and coding of 125 fake news items sourced from six fact-checking organisations in Poland. The findings revealed 6 main themes: socioeconomic, war reporting, politics, commentary, conspiracy theory and ideology. The framing of the conflict has been identified as presenting Ukrainian refugees in Poland as economic migrants, the alleged ‘Ukrainisation’ of Poland, and a narration that the war is a lie. Theoretical approach of interpreting fake news as a weapon of war revealed the pro-Russian narrative in the majority of the fake news in Poland within the analysed timeframe. The study concludes by addressing the need for a continuous research of the topic of fake news to alert the public on the tactics of the fake news spreaders and characteristics of fake news in the Polish infosphere.

The Russian Federation has recently implemented a foreign policy strategy aimed at subverting the West’s ability to deter Russia from destabilizing its neighbors. This strategy combines elements of conventional military strategy with “weaponized information” in order to achieve success in the political and military arenas of conflict. “Weaponized Information” goes beyond the “network-centric” warfare envisioned by cyber security experts, focused instead upon the development of “fake news,” disinformation, and encouraging conflicting media narratives. This thesis explores this strategy through Thomas Schelling’s framework of deterrence elucidated in Arms and Influence and uses recent events in Ukraine, Syria, the United States, and Europe to describe the development and implementation of “weaponized information” in 21st Century international conflicts. / Master of Arts

Yes / Deepfakes constitute fake content -generally in the form of video clips and other media formats such as images or audio- created using deep learning algorithms. With the rapid development of artificial intelligence (AI) technologies, the deepfake content is becoming more sophisticated, with the developed detection techniques proving to be less effective. So far, most of the detection techniques in the literature are based on AI algorithms and can be considered as passive. This paper presents a proof-of-concept deepfake detection system that detects fake news video clips generated using voice impersonation. In the proposed scheme, digital watermarks are embedded in the audio track of a video using a hybrid speech watermarking technique. This is an active approach for deepfake detection. A standalone software application can perform the detection of robust and fragile watermarks. Simulations are performed to evaluate the embedded watermark's robustness against common signal processing and video integrity attacks. As far as we know, this is one of the first few attempts to use digital watermarking for fake content detection. / EIG CONCERT-Japan call to the project entitled “Detection of fake newS on SocIal MedIa pLAtfoRms” (DISSIMILAR) through grants PCI2020-120689-2 (Ministry of Science and Innovation, Spain) and JPMJSC20C3 (JST SICORP, Japan). In addition, the work of the first two authors was partly funded by the Spanish Government through RTI2018-095094-B-C22 “CONSENT”

De plus en plus de médias dans le monde disposent de rubriques ou chroniques dédiées au fact-checking. Elles visent notamment à vérifier la véracité de propos tenus par des responsables politiques. Cette pratique revisite celle née aux États-Unis dans les années 1920, qui consistait à vérifier de manière exhaustive et systématique les contenus avant parution. Ce fact-checking « moderne » incarne une stratégie des rédactions web – en dépit des crises structurelles et conjoncturelles – pour renouer avec la diffusion de contenus mieux vérifiés, ainsi que leur capacité à mettre à profit les outils numériques qui facilitent l’accès à l’information. À travers une trentaine d’entretiens semi-directifs avec des fact-checkeurs français et l’étude de 300 articles et chroniques issus de sept médias différents, ce travail de recherche analyse dans quelle mesure le fact-checking, en tant que genre journalistique, valorise une démarche crédible, mais révèle aussi, en creux, des manquements dans les pratiques professionnelles. Il examine, enfin, comment la promotion de contenus plus qualitatifs et l’éducation aux médias sont de nature à placer le fact-checking au cœur des stratégies éditoriales, destinées à regagner la confiance des publics. / A growing number of newsrooms around the world have established fact-checking headings or rubrics. They are dedicated to assess the veracity of claims, especially by politicians. This practice revisits an older fact-checking practice, born in the United States in the 1920’s and based on an exhaustive and systematic checking of magazines’ contents before publishing. The ‘modern’ version of fact-checking embodies both the willingness of online newsrooms to restore verified contents —despite the structural and economic crisis of the press— and their ability to capitalize on digital tools which enhance access to information. Through some thirty semi-structured interviews with French fact-checkers and the study of a sample of 300 articles and chronicles from seven media, this PhD thesis examines the extent to which fact-checking, as a journalistic genre, certainly valorizes a credible method, but also —and indirectly— reveals shortcomings in professional practices. Finally, it discusses how the promotion of more qualitative content, as well as media literacy, could place fact-checking at the heart of editorial strategies —the latter aiming at retrieving trust from the audience.

Undoubtedly, the rapid development of information systems and the widespread use of electronic means and social networks have played a significant role in accelerating the pace of events worldwide, such as, in the 2012 Gaza conflict (the 8-day war), in the pro-secessionist rebellion in the 2013-2014 conflict in Eastern Ukraine, in the 2016 US Presidential elections, and in conjunction with the COVID-19 outbreak pandemic since the beginning of 2020. As the number of daily shared data grows quickly on various social networking platforms in different languages, techniques to carry out automatic classification of this huge amount of data timely and correctly are needed. Of the many social networking platforms, Twitter is of the most used ones by netizens. It allows its users to communicate, share their opinions, and express their emotions (sentiments) in the form of short blogs easily at no cost. Moreover, unlike other social networking platforms, Twitter allows research institutions to access its public and historical data, upon request and under control. Therefore, many organizations, at different levels (e.g., governmental, commercial), are seeking to benefit from the analysis and classification of the shared tweets to serve in many application domains, for examples, sentiment analysis to evaluate and determine user’s polarity from the content of their shared text, and misleading information detection to ensure the legitimacy and the credibility of the shared information. To attain this objective, one can apply numerous data representation, preprocessing, natural language processing techniques, and machine/deep learning algorithms. There are several challenges and limitations with existing approaches, including issues with the management of tweets in multiple languages, the determination of what features the feature vector should include, and the assignment of representative and descriptive weights to these features for different mining tasks. Besides, there are limitations in existing performance evaluation metrics to fully assess the developed classification systems. In this dissertation, two novel frameworks are introduced; the first is to efficiently analyze and classify bilingual (Arabic and English) textual content of social networks, while the second is for evaluating the performance of binary classification algorithms. The first framework is designed with: (1) An approach to handle Arabic and English written tweets, and can be extended to cover data written in more languages and from other social networking platforms, (2) An effective data preparation and preprocessing techniques, (3) A novel feature selection technique that allows utilizing different types of features (content-dependent, context-dependent, and domain-dependent), in addition to (4) A novel feature extraction technique to assign weights to the linguistic features based on how representative they are in in the classes they belong to. The proposed framework is employed in performing sentiment analysis and misleading information detection. The performance of this framework is compared to state-of-the-art classification approaches utilizing 11 benchmark datasets comprising both Arabic and English textual content, demonstrating considerable improvement over all other performance evaluation metrics. Then, this framework is utilized in a real-life case study to detect misleading information surrounding the spread of COVID-19. In the second framework, a new multidimensional classification assessment score (MCAS) is introduced. MCAS can determine how good the classification algorithm is when dealing with binary classification problems. It takes into consideration the effect of misclassification errors on the probability of correct detection of instances from both classes. Moreover, it should be valid regardless of the size of the dataset and whether the dataset has a balanced or unbalanced distribution of its instances over the classes. An empirical and practical analysis is conducted on both synthetic and real-life datasets to compare the comportment of the proposed metric against those commonly used. The analysis reveals that the new measure can distinguish the performance of different classification techniques. Furthermore, it allows performing a class-based assessment of classification algorithms, to assess the ability of the classification algorithm when dealing with data from each class separately. This is useful if one of the classifying instances from one class is more important than instances from the other class, such as in COVID-19 testing where the detection of positive patients is much more important than negative ones. / Graduate

[pt] A credibilidade está em xeque de uma maneira global, o que ameaça a autoridade de instituições em geral, incluindo a imprensa, na chamada era da pósverdade. Por tratar com desimportância a verdade, os tempos atuais têm sido marcados pela circulação de informações falsas. É um tempo também conhecido como era pós-factual. A mesma sociedade em rede impulsionada pela internet com promessas de futuro mais horizontalizado viu nascer a plataformização algorítmica, a desqualificação da imprensa, a cultura da desinformação. É desse ambiente extremamente hostil e vulnerável para as instituições democráticas que partem as reflexões trazidas nesta tese. Tamanho é o nível de desinformação que hoje há uma proliferação de agências e serviços de checagem em diversos países. Por princípio, o fact-checking existe para apontar se discursos são verdadeiros ou falsos. No entanto, essas agências e serviços ganharam visibilidade e tornaram-se mais influentes no mercado jornalístico. O deslocamento do fact-checking para ambientes externos às redações de veículos de imprensa ou para espaços específicos nos veículos traz consigo uma pressuposição de que possa haver um deslocamento de uma das funções mais importantes do jornalismo: a checagem, que contribui para a construção de credibilidade e, consequentemente, funciona como sustentáculo da autoridade jornalística. Assim, o objetivo geral aqui foi entender qual o lugar ocupado pelas agências de checagem no universo do jornalismo brasileiro e que autoridade jornalística elas têm reivindicado. Uma questão central foi de que maneira o trabalho de fact-checking influencia a credibilidade/autoridade jornalística e se há – e em que medida – comprometimento dessa autoridade a partir do estabelecimento mercadológico dos serviços de checagem. Para isso, analisamos o conteúdo bruto de 677 posts sobre as eleições para presidência do Brasil feitos pelas agências Lupa e Aos Fatos e pelos serviços Fato ou Fake, Estadão Verifica e Folha Informações no período entre 16 de agosto e 30 de outubro de 2022, que compreende o dia do início da campanha eleitoral e o dia da votação do segundo turno das eleições. / [en] Credibility is being called into question in a global way, which threatens the authority of institutions in general, including the press, in the so-called post-truth era;. Due to its disregard for truth, the current times have been marked by the circulation of fake news. It is also a time known as the post-factual era. The same networked society driven by the internet with promises of a more horizontal future has witnessed the rise of algorithmic platformization, the disqualification of the press, and the culture of misinformation. It is from this extremely hostile and vulnerable environment for democratic institutions that the reflections brought in this thesis arise. The level of misinformation is such that today there is a proliferation of fact-checking agencies and services in various countries. In principle, fact-checking exists to determine whether speeches are true or false. However, these agencies and services have gained visibility and become more influential in the journalistic environment. The shift of fact-checking to external environments outside newsrooms or to specific spaces within news outlets carries with it an assumption that one of the most important functions of journalism, factchecking, can be displaced. This function contributes to the construction of credibility and, consequently, functions as the pillar of journalistic authority. Thus, the general objective here was to understand the place occupied by fact-checking agencies in the Brazilian journalism universe and the journalistic authority they have claimed. A central issue is how fact-checking work influences journalistic credibility/authority and whether there is - and to what extent – a compromise of this authority through the market establishment of fact-checking services. To this end, we analyzed the raw content of 677 posts about the presidential elections in Brazil made by the Lupa and Aos Fatos agencies and the Fato ou Fake, Estadão Verifica, and Folha Informações services in the period between August 16 and October 30, 2022, which includes the day campaign started and the day of the second round of voting in the elections.

Vårt samhälle blir alltmer digitaliserat och för generationer födda innan internets intågande i hemmen har inte användandet av internet och sociala medier alltid varit en självklarhet. Därför undersöker denna studie hur sociala och personliga aspekter hos svenska personer över 70 år påverkar motivation och attityd i användandet av sociala medier, hur de upplever sin digitala kunskap, hur de upplever att de förhåller sig källkritiskt till information på sociala medier och om de sprider informationen vidare. Studien använde sig av teorier om digitala klyftor, mediekompetens, falsk information och källkritik och genomfördes genom kvalitativa intervjuer. Resultatet visade att sociala aspekter har stor betydelse för deltagarnas motivation och attityd i användandet av sociala medier och att deltagarna upplever brister i sin digitala kunskap, men att få ville vidareutveckla den. De flesta upplever att de tänker källkritiskt kring information som de anser är av viktigt och om den inte bedöms vara viktig kontrolleras den inte. Deltagarna i denna studie anser sig inte sprida information från sociala medier i någon större utsträckning. / Our society is becoming increasingly digital and for generations born before the Internet became an integrated part of peoples home the use of the internet and social media has not always been an obvious part of daily life. Therefore, this study examines how social and personal aspects affect motivation and attitude in the use of social media in Swedish people over the age of 70, how they experience their digital literacy, how source critical they experience themselves to be towards information on social media and whether they are spreading the information to others. The study used theories of digital divide, media literacy, misinformation and source criticism and was conducted through qualitative interviews. The results showed that social aspects are of great importance for the participants' motivation and attitude in the use of social media and that the participants experienced shortcomings in their digital literacy, but few wanted to further develop it. Most of them experience that they think critically about information that they consider to be important and if it is not deemed to be important, it is not checked. The participants in this study do not consider themselves disseminating information from social media to any great extent.

In today’s world, where online news is so widespread, various methods have been developed in order to provide users with personalized news recommendations. Wonderful accomplish ments have been made when it comes to providing readers with everything that could attract their attention. While accuracy is critical in news recommendation, other factors, such as diversity, novelty, and reliability, are essential in satisfying the readers’ satisfaction. In fact, technological advancements bring additional challenges which might have a detrimental im pact on the news domain. Therefore, researchers need to consider the new threats in the development of news recommendations. Fake news, in particular, is a hot topic in the media today and a new threat to public safety. This work presents a modularized system capable of recommending news to the user and detecting fake news, all while helping users become more aware of this issue. First, we suggest FANAR, FAke News Aware Recommender system, a modification to news recommendation algorithms that removes untrustworthy persons from the candidate user’s neighbourhood. To do this, we created a probabilistic model, the Beta Trust model, to calculate user rep utation. For the recommendation process, we employed Graph Neural Networks. Then, we propose EXMULF, EXplainable MUltimodal Content-based Fake News Detection Sys tem. It is tasked with the veracity analysis of information based on its textual content and the associated image, together with an Explainable AI (XAI) assistant that is tasked with combating the spread of fake news. Finally, we try to raise awareness about fake news by providing personalized alerts based on user reliability. To fulfill the objective of this work, we build a new dataset named FNEWR. Our exper iments reveal that EXMULF outperforms 10 state-of-the-art fake news detection models in terms of accuracy. It is also worth mentioning that FANAR , which takes into account vi sual information in news, outperforms competing approaches based only on textual content. Furthermore, it reduces the amount of fake news found in the recommendations list / De nos jours, où les actualités en ligne sont si répandues, diverses méthodes ont été dé veloppées afin de fournir aux utilisateurs des recommandations d’actualités personnalisées. De merveilleuses réalisations ont été faites lorsqu’il s’agit de fournir aux lecteurs tout ce qui pourrait attirer leur attention. Bien que la précision soit essentielle dans la recommandation d’actualités, d’autres facteurs, tels que la diversité, la nouveauté et la fiabilité, sont essentiels pour satisfaire la satisfaction des lecteurs. En fait, les progrès technologiques apportent des défis supplémentaires qui pourraient avoir un impact négatif sur le domaine de l’information. Par conséquent, les chercheurs doivent tenir compte des nouvelles menaces lors de l’élabo ration de nouvelles recommandations. Les fausses nouvelles, en particulier, sont un sujet brûlant dans les médias aujourd’hui et une nouvelle menace pour la sécurité publique. Au vu des faits mentionnés ci-dessus, ce travail présente un système modulaire capable de détecter les fausses nouvelles, de recommander des nouvelles à l’utilisateur et de les aider à être plus conscients de ce problème. Tout d’abord, nous suggérons FANAR, FAke News Aware Recommender system, une modification d’algorithme de recommandation d’actuali tés qui élimine les personnes non fiables du voisinage de l’utilisateur candidat. A cette fin, nous avons créé un modèle probabiliste, Beta Trust Model, pour calculer la réputation des utilisateurs. Pour le processus de recommandation, nous avons utilisé Graph Neural Net works. Ensuite, nous proposons EXMULF, EXplainable MUltimodal Content-based Fake News Detection System. Il s’agit de l’analyse de la véracité de l’information basée sur son contenu textuel et l’image associée, ainsi qu’un assistant d’intelligence artificielle Explicable (XAI) pour lutter contre la diffusion de fake news. Enfin, nous essayons de sensibiliser aux fake news en fournissant des alertes personnalisées basées sur le profil des utilisateurs. Pour remplir l’objectif de ce travail, nous construisons un nouveau jeu de données nommé FNEWR. Nos résultats expérimentaux montrent qu’EXMULF surpasse 10 modèles de pointe de détection de fausses nouvelles en termes de précision. Aussi, FANAR qui prend en compte les informations visuelles dans les actualités, surpasse les approches concurrentes basées uniquement sur le contenu textuel. De plus, il permet de réduire le nombre de fausses nouvelles dans la liste des recommandations.

[pt] Nos últimos anos, países democráticos presenciaram a participação de fake news em seus processos deliberativos. A forma como ela ocorreu variou de acordo com as especificidades da estrutura comunicacional de cada nação. A expressiva presença de fake news em eleições recentes suscitou a preocupação sobre a possibilidade de esses materiais estarem interferindo em processos deliberativos, causando contextos de desinformação, enganando e manipulando eleitores e, consequentemente, afetando seus resultados. Esta dissertação analisa a participação das fake news na eleição presidencial de 2016 nos Estados Unidos, bem como nas eleições presidenciais de 2018 e municipais de 2020 no Brasil. Busca compreender de que forma esses conteúdos foram utilizados em cada um dos contextos acima, quais os mecanismos de disseminação usados, as abordagens adotadas, os principais temas levantados, a abrangência da sua difusão, as possíveis influências que podem ter exercido e a sua relevância em cada um desses contextos. / [en] In recent years, democratic countries have witnessed the participation of fake news in their deliberative processes. The way it has taken place has varied according to the specifics of the communication structure of each nation. The significant presence of fake news in recent elections has raised concerns about the possibility that these materials are interfering with deliberative processes, causing contexts of disinformation, deception and manipulation of voters, consequently affecting their results. This dissertation analyzes the participation of fake news in the 2016 presidential election in the United States, as well as in the 2018 presidential and 2020 municipal elections in Brazil. It seeks to understand how these contents were used in each of the above contexts, which dissemination mechanisms were used, the approaches adopted, the main issues raised, the extension of their diffusion, the possible influences they may have had and their relevance in each one of those contexts.

Fake news, or information disorder, is a societal problem that could be partially remedied by automatic detection tools. While still a young research field many such tools have been proposed in academic writing. This systematic mapping study gives an overview of the current research in Natural Language Process-based fake news detection utilising Machine Learning and Neural Network classification algorithms in regards to which classification algorithms have been studied and which datasets have been used. Furthermore, we attempt to make a generalised description of the performance (measured in f-score and accuracy) of the most commonly occurring classification algorithms. From a corpus of 124 research articles and other scientific texts we identify 63 different datasets mainly written in English, and 116 different classification algorithms. The seven most commonly occurring algorithms (Random Forest, Logistic Regression, Support Vector Machine, Decision Tree, Long Short- TermMemory, K-Nearest Neighbors, Convolutional Neural Network) together make up almost 50% of all algorithm occurences in the article corpus. For these seven, the ten occurrences with the best performance are listed. Out of the datasets, the six most common datasets (ISOT, FakeNewsNet, Patwa 2021, LIAR, Bisaillon, and UTK-MLC) together make up 44% of all dataset occurrences. Apart from English, the represented languages were mainly Chinese (Mandarin), Portugese, Indonesian, Bangla, and Albanian. / Olika typer av desinformation (så kallade fake news), är ett problem för dagens samhälle. En av flera möjliga dellösningar på problemet utgörs av automatiserad fake news-detektion. Trots att detta forskningsfält är relativt nytt finns det en uppsjö av olika föreslagna modeller för automatiserad fake news-detektion. Denna systematiska kartläggning syftar till att ge en överblick över den aktuella forskningen inom Natural Language Processing-baserad automatiserad fake news-detektion med klassifikationsalgoritmer både inom maskininlärning och neurala nätverk. Översikten avser vilka klassifikationsalgoritmer samt vilka dataset som förekommer inom forskningen. Vidare försöker vi göra en generell beskrivning av prestandan hos de vanligast förekommande klassifikationsalgoritmerna, mätt i accuracy och f-score. Kartläggningen omfattar en samling på 124 artiklar och andra vetenskapliga texter, ur vilka vi identifierade 63 förekommance dataset och 116 olika förekommande klassifikationsalgoritmer. De sju vanligast förekommande algoritmerna (Random Forest, Logistic Regression, Support Vector Machine, Decision Tree, Long-Short Memory Network, K-Nearest Neighbors, Convolutional Neural Network) utgör tillsammans 49% av alla förekomster inom artikelsamlingen. Vi har tagit ut santliga förekomster av prestandaresultat för dessa sju algoritmer, och listat de tio bästa prestandaresultaten för var och en av de sju algoritmerna. De sex vanligast förekommande dataseten (ISOT, FakeNewsNet, Patwa 2021, LIAR, Bisaillon, and UTK-MLC) utgör tillsammans 44% av alla förekomster. Engelska var med stor marginal det vanligast förekommande språket inom dataseten, andra språk som förekom var kinesiska (mandarin), portugisiska, indonesiska, bangla, och albanska.