Computational Data and Social Networks: 10th International Conference, CSoNet 2021, Virtual Event, November 15–17, 2021, Proceedings (Theoretical Computer Science and General Issues)

This book constitutes the refereed proceedings of the 10th International Conference on Computational Data and Social Networks, CSoNet 2021, which was held online during November 15-17, 2021. The conference was initially planned to take place in Montreal, Quebec, Canada, but changed to an online event due to the COVID-19 pandemic. The 24 full and 8 short papers included in this book were carefully reviewed and selected from 57 submissions. They were organized in topical sections as follows: Combinatorial optimization and learning; deep learning and applications to complex and social systems; measurements of insight from data; complex networks analytics; special track on fact-checking, fake news and malware detection in online social networks; and special track on information spread in social and data networks. Preface Organization Contents Combinatorial Optimization and Learning Streaming Algorithms for Maximizing Non-submodular Functions on the Integer Lattice 1 Introduction 1.1 Additional Related Work 2 Preliminaries 3 Streaming Algorithm with Known OPT 4 Streaming Algorithm with Known f(Xe) 5 The One Pass Streaming Algorithm References Causal Inference for Influence Propagation—Identifiability of the Independent Cascade Model 1 Introduction 2 Related Work 3 Model and Problem Definitions 4 Parameter Identifiability of the Markovian IC Model 5 Parameter Identifiability of the Semi-Markovian IC Model 5.1 Condition on Unidentifiability of the Semi-Markovian IC Model 5.2 Identifiability of the Chain Model 6 Parameter Identifiability of Model with a Global Hidden Variable 6.1 Observable IC Model with only a Global Hidden Variable 6.2 Markovian IC Model with a Global Hidden Variable (Mixed Model) 7 Conclusion References Streaming Algorithms for Budgeted k-Submodular Maximization Problem 1 Introduction 2 Preliminaries 3 Deterministic Streaming Algorithm When =1 3.1 Deterministic Streaming Algorithm with Known Optimal Solution 3.2 Deterministic Streaming Algorithm 4 Random Streaming Algorithm for General Case 4.1 Random Algorithm with Known Optimal Solution 4.2 Random Streaming Algorithm 5 Conclusions References Approximation Algorithms for the Lower Bounded Correlation Clustering Problem 1 Introduction 2 Lower Bounded Correlation Clustering Problem 3 A Simple Effecient Algorithm 3.1 Avgcen(V)(V)17/80 3.2 Avgcen(V)(V)>17/80 4 A Complex Algorithm May Outputs Multiple Clusters 4.1 Disagreements Generated by Positive Edges 4.2 Disagreements Generated by Negative Edges 5 Discussions References Approximation Algorithm for Maximizing Nonnegative Weakly Monotonic Set Functions 1 Introduction 2 Preliminaries 3 The Deterministic-Greedy Algorithm 4 Discussion References Differentially Private Submodular Maximization over Integer Lattice 1 Introduction 1.1 Our Contributions 1.2 Related Work 1.3 Organizations 2 Preliminaries 2.1 Composition of Differential Privacy 2.2 Exponential Mechanism 3 Algorithm 3.1 Diff. Private DR-Submodular Algorithm 3.2 Diff. Private Integer Submodular Algorithm 4 Conclusion References Maximizing the Sum of a Supermodular Function and a Monotone DR-submodular Function Subject to a Knapsack Constraint on the Integer Lattice 1 Introduction 2 Preliminaries 3 The Streaming Algorithms 3.1 Algorithms with Known Optimum 3.2 Algorithms with Unknown Optimum 4 Conclusions References Deep Learning and Applications to Complex and Social Systems A Framework for Accelerating Graph Convolutional Networks on Massive Datasets 1 Introduction 2 Technical Details 2.1 Background 2.2 Existing Methods, Memory Requirements, and Data Loading Costs 3 Overall Approach and Implementations 3.1 Background: Meta-learning Approach 3.2 Our Approach 4 Experimental Results 4.1 Implementations and Setup 4.2 Datasets 4.3 F1-Score 4.4 Training Times 5 Related Work 6 Conclusions References AdvEdge: Optimizing Adversarial Perturbations Against Interpretable Deep Learning 1 Introduction 2 Related Work 3 Fundamental Concepts 4 AdvEdge Attack 4.1 Attack Definition 4.2 Interpretation Models 5 Experimental Setting 6 Attack Evaluation 6.1 Attack Effectiveness Against DNNs 6.2 Attack Effectiveness Against Interpreters 6.3 Adversarial Perturbation Rate 7 Conclusion References Incorporating Transformer Models for Sentiment Analysis and News Classification in Khmer 1 Introduction 2 Background Study 3 Data Description 3.1 Dataset for BERT Pre-training 3.2 Dataset for BERT Fine-Tuning 4 Data Preprocessing 4.1 Basic Preprocessing 4.2 Word Segmentation 5 Methodology 5.1 FastText 5.2 BERT 6 Results and Analysis 6.1 Sentiment Analysis 6.2 News Classification 7 Conclusion References Deep Bangla Authorship Attribution Using Transformer Models 1 Introduction 2 Related Works 3 Data Description 4 Methodology 4.1 Data Preprocessing 4.2 Pre-trained Transformer Models 4.3 Fine-Tuning Transformer Models 4.4 Evaluation Metrics 5 Result and Analysis 6 Conclusion and Future Work References A Deep Learning Based Traffic Sign Detection for Intelligent Transportation Systems 1 Introduction 2 Proposed System Architecture 3 Experiment 3.1 Datasets 3.2 Data Preprocessing 3.3 Evaluation Methods 3.4 Results 3.5 SIGN Detection Application 3.6 Error Analysis 4 Concluding Remarks References Detecting Hate Speech Contents Using Embedding Models 1 Introduction 2 Related Work 3 Proposed Method 3.1 Input Representation 3.2 Neural Network Models 4 Experiments 4.1 Datasets 4.2 Experimental Results 5 Conclusion References MIC Model for Cervical Cancer Risk Factors Deep Association Analysis 1 Introduction 2 MIC Model 2.1 Multiple Indicators Correlation Model 3 The Experimental Analysis 4 Conclusion References Power Grid Cascading Failure Prediction Based on Transformer 1 Introduction 2 Literature Review 3 Models 3.1 Cascading Failure Model 3.2 Transformer Model 3.3 Independent Cascade Model 4 Experiments 4.1 Transformer Model Hyperparameter 4.2 Total Number of Failed Lines Prediction 4.3 Line Failure Frequency 4.4 Line Failure Magnitude 4.5 F1 Score 4.6 IC Model Simulation 4.7 Computational Efficiency 5 Conclusion References Measurements of Insight from Data Security Breaches in the Healthcare Domain: A Spatiotemporal Analysis 1 Introduction 2 Data Sources 3 Studied Dimensions and Variables 4 Measurement Results and Discussions 4.1 The Global Distribution of Incidents 4.2 Number of Incidents by State 4.3 Analyzing the Compromised Assets 4.4 State Level Correlation 4.5 Organizations Size 4.6 Timeline Discovery 4.7 Internal and External Discovery Methods 4.8 Targeted vs Opportunistic 5 Analysis of the OCR Dataset 6 Related Work 7 Conclusion References Social and Motivational Factors for the Spread of Physical Activities in a Health Social Network 1 Introduction 2 YesiWell Study 3 Self-motivation, Influence, and Susceptibility 4 Experimental Results 5 Conclusions References Understanding the Issues Surrounding COVID-19 Vaccine Roll Out via User Tweets 1 Introduction 2 Data and Methods 3 Linguistic Analysis 3.1 Term Frequencies—Unigrams and Bigrams 3.2 Topic Modeling 4 Sentiment Analysis 5 Related Work 6 Conclusions References Complex Networks Analytics Minimize Travel Time with Traffic Flow Density Equilibrium on Road Network 1 Introduction 2 Road Network 3 Problem Formulation and Algorithm Design 3.1 Formula Description 3.2 Algorithm Design 4 Simulation Result 4.1 Simulation Setup 4.2 Simulation Analysis 5 Conclusion References Network Based Framework to Compare Vaccination Strategies 1 Introduction 2 Related Work 3 Network Based Framework 4 Considered Vaccination Strategies 5 Results and Discussion 5.1 Comparison with the Real-World Data 5.2 Disease Spread 5.3 Comparison of Vaccination Strategies and Effect of Starting Date of Vaccination 6 Conclusion and Future Directions References Groups Influence with Minimum Cost in Social Networks 1 Introduction 2 Related Work 3 Diffusion Models and Problem Definition 3.1 Independent Cascade Model 3.2 Problem Definition 3.3 An Estimator of Group Influence 4 Proposed Algorithm 5 Experiments 5.1 Experimental Settings 5.2 Experiment Results 6 Conclusion References Recovering Communities in Temporal Networks Using Persistent Edges 1 Introduction 2 Degree-Corrected Temporal Network Model with Markov Edge Dynamics 2.1 Model Description 2.2 Maximum Likelihood Estimator 2.3 Temporal Spectral Clustering Combining New and Persistent Edges 3 Numerical Experiments 3.1 Synthetic Data 3.2 Social Networks of High School Students A Proofs of Main Statements A.1 Maximum Likelihood Computations (Proposition 1) A.2 Approximation of the MLE A.3 Modularity and Normalized Spectral Clustering References Community Detection Using Semilocal Topological Features and Label Propagation Algorithm 1 Introduction 2 Proposed Method 2.1 Time Complexity 3 Experimental Analysis 3.1 Datasets 3.2 Experimental Settings 3.3 Performance Analysis 3.4 Sensitivity Analysis 4 Conclusion References Twitter Analysis of Covid-19 Misinformation in Spain 1 Introduction 2 Dataset and Methodology 3 Tweets Analysis 3.1 Graphs Analysis 3.2 Tweets Influence 4 User Analysis 4.1 User Graph 4.2 User Influence 4.3 Users Location 5 Discussion and Future Work References Comparing Community-Aware Centrality Measures in Online Social Networks 1 Introduction 2 Community-Aware Centrality Measures 3 Data, Tools, And Methods 3.1 Data 3.2 Susceptible-Infected-Removed Model 3.3 Imprecision Function 3.4 Methods 4 Experimental Results 4.1 Performance of the community-Aware Centrality Measures Within Networks 4.2 Performance of the community-Aware Centrality Measures Across Networks 4.3 Influence of the transmission Rate 5 Discussion 6 Conclusion References Two-Tier Cache-Aided Full-Duplex Content Delivery in Satellite–Terrestrial Networks 1 Introduction 2 System Model 2.1 Two-Tier Caching Model 2.2 Channel Model 2.3 Transmission Scheme 3 Successful Delivery Probability 4 Satellite Bandwidth Consumption 5 Cache Placement Design 6 Numerical Results 7 Conclusion References Special Track: Fact-Checking, Fake News and Malware Detection in Online Social Networks Mean User-Text Agglomeration (MUTA): Practical User Representation and Visualization for Detection of Online Influence Operations 1 Introduction 2 Ethical Considerations and Social Impact 3 Related Work 4 Methodology 4.1 User Representation 5 Datasets and Processing 6 Experiments and Evaluation 6.1 Assessment of Visualization Quality 6.2 Qualitative Analysis and Application 7 Conclusion References The Role of Information Organization and Knowledge Structuring in Combatting Misinformation: A Literary Analysis 1 Introduction 2 Information Access as a Factor 2.1 Access as a Definitional Construct 2.2 Applications on Misinformation 3 Information Discovery as an Amplifier 3.1 Discovery as a Definitional Construct 3.2 Applications on Misinformation 4 Information Retrieval as the Filter of Fake News 4.1 Retrieval as a Definitional Construct 4.2 Applications on Misinformation 5 The Contribution of Folksonomy in Misinformation Explosion 5.1 Influence in Social Networking Platforms 5.2 The Rise of Misinformation 5.3 Crowd-Driven Tags as Enforcers of Misinformation 5.4 Online Communities of Misinformation 6 The Classical Functions of Taxonomic Authority as Misinformation Safeguards 6.1 Natural Characteristics of Taxonomy Against Misinformation 6.2 The Bias Neutralizer in News and Media 6.3 Achieving Conjunctive Balance with Folksonomy 7 Conclusion 8 Limitations References Fake News Detection Using LDA Topic Modelling and K-Nearest Neighbor Classifier 1 Introduction 2 Background 3 The Proposed Approach 3.1 Feature Extraction 3.2 News Classification 4 Experimental Results 5 Conclusions References Machine Learning Technique for Fake News Detection Using Text-Based Word Vector Representation 1 Introduction 2 Related Work 3 Methodology 3.1 Dataset 3.2 Pre-processing of Dataset 3.3 Machine Learning Models 4 Result and Discussions 4.1 Confusion Matrix Calculation 4.2 Performance Comparison 5 Conclusion References Special Track: Information Spread in Social and Data Networks Summarization Algorithms for News: A Study of the Coronavirus Theme and Its Impact on the News Extracting Algorithm 1 Introduction 2 Research Method 3 Data Sources and Modelling 4 Results and Discussion 5 Conclusion Appendix References Social Cohesion During the Stay-at-Home Phase of the First Wave of the COVID-19 Pandemic on Polish-Speaking Twitter 1 Introduction 2 Data, Methods and Research Questions 3 Results 4 Discussion 5 Conclusion References Target Set Selection in Social Networks with Influence and Activation Thresholds 1 Introduction 1.1 Motivation and Problem Description 1.2 Literature Review 1.3 Contribution and Organization 2 Minimum Target Set Selection Model with Influence and Activation Thresholds 3 Computational Algorithms 3.1 Graph Partition 3.2 Nodes Selection 3.3 Greedy Algorithm 4 Computational Algorithms Comparison 5 Conclusion References Extended Abstracts Social Activity and Decentralized Applications in Blockchain-Based Social Networks References Vulnerabilities Assessment of Deep Learning-Based Fake News Checker Under Poisoning Attacks 1 Introduction 2 The Envised Case Study 3 Conclusion and Future Works Transfer Learning and Loan Default Prediction 1 Introduction 2 Datasets and Methods References Author Index