Data Science for Fake News: Surveys and Perspectives (The Information Retrieval Series, 42)

This book provides an overview of fake news detection, both through a variety of tutorial-style survey articles that capture advancements in the field from various facets and in a somewhat unique direction through expert perspectives from various disciplines. The approach is based on the idea that advancing the frontier on data science approaches for fake news is an interdisciplinary effort, and that perspectives from domain experts are crucial to shape the next generation of methods and tools. The fake news challenge cuts across a number of data science subfields such as graph analytics, mining of spatio-temporal data, information retrieval, natural language processing, computer vision and image processing, to name a few. This book will present a number of tutorial-style surveys that summarize a range of recent work in the field. In a unique feature, this book includes perspective notes from experts in disciplines such as linguistics, anthropology, medicine and politics that will help to shape the next generation of data science research in fake news. The main target groups of this book are academic and industrial researchers working in the area of data science, and with interests in devising and applying data science technologies for fake news detection. For young researchers such as PhD students, a review of data science work on fake news is provided, equipping them with enough know-how to start engaging in research within the area. For experienced researchers, the detailed descriptions of approaches will enable them to take seasoned choices in identifying promising directions for future research. Preface......Page 6 Acknowledgments......Page 8 Contents......Page 9 1 Introduction......Page 15 2 Surveys......Page 17 2.3 Deep Learning Methods for Fake News Detection......Page 19 2.5 Neural Text Generation......Page 20 2.7 Graph Mining Meets Fake News Detection......Page 21 3 Perspectives......Page 22 3.1 Fake News in Health Sciences......Page 23 3.3 A Political Science Perspective on Fake News......Page 24 3.5 Misinformation and the Indian Election......Page 25 4 Concluding Remarks......Page 26 References......Page 27 Part I Survey......Page 28 1 Introduction......Page 29 1.1 Paradigms of Machine Learning vis-à-vis Supervision......Page 30 1.2 Challenges for Unsupervised Learning in Fake News Detection......Page 31 2.1 Conceptual Basis for UFND Methods......Page 33 Differentiating User Types......Page 35 Propagandist Patterns......Page 36 2.3 Building Blocks for UFND......Page 38 3.1 Truth Discovery......Page 40 3.2 Differentiating User Types......Page 42 3.3 Propagandist Patterns......Page 44 3.4 Inter-user Dynamics......Page 45 4.1 Specialist Domains and Authoritative Sources......Page 47 4.2 Statistical Data for Fake News Detection......Page 48 4.3 Early Detection......Page 49 5 Conclusions......Page 50 References......Page 51 1 Introduction......Page 53 2 Challenges and Opportunities......Page 56 3 Multi-modal Fake News Datasets......Page 57 3.1 Fake Microblog Datasets......Page 60 3.2 Fake News Datasets......Page 61 4 State-of-the-Art Models......Page 62 5 Unsupervised Approach......Page 63 6.1 JIN......Page 66 6.2 TI-CNN......Page 67 6.3 MKEMN......Page 68 6.4 SpotFake and SpotFake+......Page 69 6.6 SAFE......Page 70 7.1 AGARWAL......Page 71 7.2 MVNN......Page 72 8 Hybrid Fusion Approach......Page 74 9 Adversarial Model......Page 75 9.1 SAME......Page 76 10 Autoencoder Model......Page 77 References......Page 78 1 Introduction......Page 83 1.1 Fake News Types......Page 84 1.2 Early Works......Page 85 2 Deep Learning Methods......Page 86 2.1 Fake News Detection Using CNN......Page 88 2.2 Fake News Detection Using RNN and Its Variants......Page 90 2.3 Multimodal Methods......Page 97 3.1 Datasets......Page 99 3.2 Evaluation Metrics......Page 101 4 Trends in Fake News Detection Using Deep Learning......Page 102 4.1 Geometric Deep Learning......Page 103 4.2 Explainable Fake News Detection......Page 104 4.4 Neural Fake News Detection......Page 105 4.5 Discussion......Page 106 References......Page 107 1 Introduction......Page 113 3 Fake News Diffusion on Twitter......Page 115 4 Role of Bots in Spreading Fake News......Page 118 6 Identifying the Sources of Fake News......Page 119 7.1 Susceptible-Infected-Recovered (SIR) Model......Page 121 Network Construction......Page 122 Component I: Diffusion Dynamics......Page 123 Component II: Updating Personal Belief......Page 124 7.3 Percolation Model......Page 125 Branching and Size of Cascade......Page 126 7.4 Spread-Stifle Model......Page 127 How the Spread-Stifle Model Differs from Others?......Page 128 Mean-Field Approach......Page 129 Reachability Probabilities......Page 130 Transition Probabilities......Page 131 Mean-Field Rate of Change......Page 133 8 Strategies to Minimize the Spread of Fake News......Page 135 9 Summary of the Chapter......Page 136 References......Page 138 1 Introduction......Page 140 2 Modeling Approaches......Page 141 2.2 Language Models......Page 142 2.3 Encoder–Decoder Attention......Page 143 2.5 Seq2Seq Model......Page 144 3.1 Supervised Learning Techniques......Page 145 3.4 Embedding Techniques......Page 146 4.1 Contextualized Word Vectors (CoVe)......Page 147 4.3 BERT......Page 148 4.5 Transformer-XL......Page 149 4.6 Larger Language Models......Page 150 4.8 GROVER......Page 151 4.9 CTRL......Page 152 4.11 Discussion......Page 153 5 (Fake?) News Generation and Future Prospects......Page 154 6 Conclusion......Page 155 References......Page 156 1 Introduction......Page 159 2.1 Knowledge Graph......Page 160 2.2 RDF......Page 161 4 Knowledge Linker......Page 162 5 PredPath......Page 166 6 Knowledge Stream......Page 172 References......Page 176 Graph Mining Meets Fake News Detection......Page 179 1 Characteristics and Challenges......Page 180 2.1 Information......Page 181 2.2 Graph Models......Page 182 3.1 Graph Statistics Detection......Page 183 3.2 Dense Subgraph Mining......Page 186 3.3 Benefits and Issues......Page 188 4 Multi-modal Scenario......Page 189 4.1 Dynamic Graph-Based Approaches......Page 190 4.2 Graph-Assisted Learning Approaches......Page 192 5 Summary of the Chapter......Page 196 References......Page 197 Part II Perspectives......Page 200 Fake News in Health and Medicine......Page 201 2 Stanford University Study: Cannabis, a Cure for Cancer......Page 203 3 NBC News Study......Page 204 5 Polarised Facts......Page 205 6 Fake News During the Pandemic......Page 206 7 Consequences of Health Misinformation......Page 208 8 Managing Health Misinformation......Page 210 References......Page 211 1 Introduction......Page 213 2 Ethical Dimensions of DFND......Page 214 2.1 Mismatch of Values......Page 216 2.2 Nature of Data-Driven Learning......Page 220 2.3 Domain Properties......Page 223 3 Fairness and DFND......Page 225 4 Democratic Values and Uptake of DFND......Page 230 5 Conclusions......Page 238 References......Page 239 1 Introduction......Page 241 2 The Origins of Fake News......Page 244 3 Fake News in the Twenty-First Century......Page 245 4 Fake News and the Study of Politics......Page 247 References......Page 249 1 Introduction......Page 252 2 Sociological Studies of Disinformation......Page 255 3 Vaccine Hesitancy......Page 256 4 Elections......Page 258 5 Other Social Processes......Page 260 6 Conclusions......Page 261 References......Page 262 Misinformation and the Indian Election: Case Study......Page 264 1.1 Misinformation and Disinformation in India......Page 265 1.2 Closed Networks for Disinformation......Page 266 1.3 Scale, Prevalence, and Complexity of the Problem......Page 267 1.4 Early Solutions and Fact-Checking in India......Page 268 2.1 Automation to Augment Value......Page 269 2.2 Credibility vs. Veracity......Page 270 3.1 Credibility Assessment......Page 271 Content Analysis......Page 272 Findings During Indian Elections......Page 274 Credibility Assessment: Evaluation......Page 275 4.1 Methodology: The Life Cycle of a Claim......Page 276 4.2 Methodology During Indian Elections......Page 277 Findings During Indian Elections......Page 278 5 WhatsApp Solution for a Sharing Nation......Page 280 5.2 Related Work......Page 282 5.3 Exposing Misinformation on Closed Networks......Page 283 5.4 Disseminating Verifications to Audiences Exposed to Mis/Disinformation......Page 285 STS, Data Science, and Fake News: Questions and Challenges......Page 288 2 Truth, Power, and Knowledge......Page 289 3 Truth Versus Post-truth......Page 291 References......Page 292 1 Introduction......Page 293 1.1 Defining Fake News......Page 294 1.2 Linguistics, Sub-disciplines, and Methods......Page 295 1.3 News in Linguistics......Page 297 1.5 Different Texts and Contexts......Page 298 2.1 Bag of Words and LIWC......Page 300 2.2 Readability and Punctuation......Page 302 3 Conclusions......Page 303 References......Page 306