Ebook: Social Network Data Analytics

Social network analysis applications have experienced tremendous advances within the last few years due in part to increasing trends towards users interacting with each other on the internet. Social networks are organized as graphs, and the data on social networks takes on the form of massive streams, which are mined for a variety of purposes.

Social Network Data Analytics covers an important niche in the social network analytics field. This edited volume, contributed by prominent researchers in this field, presents a wide selection of topics on social network data mining such as Structural Properties of Social Networks, Algorithms for Structural Discovery of Social Networks and Content Analysis in Social Networks. This book is also unique in focussing on the data analytical aspects of social networks in the internet scenario, rather than the traditional sociology-driven emphasis prevalent in the existing books, which do not focus on the unique data-intensive characteristics of online social networks. Emphasis is placed on simplifying the content so that students and practitioners benefit from this book.

This book targets advanced level students and researchers concentrating on computer science as a secondary text or reference book. Data mining, database, information security, electronic commerce and machine learning professionals will find this book a valuable asset, as well as primary associations such as ACM, IEEE and Management Science.

---

Social network analysis applications have experienced tremendous advances within the last few years due in part to increasing trends towards users interacting with each other on the internet. Social networks are organized as graphs, and the data on social networks takes on the form of massive streams, which are mined for a variety of purposes. Social Network Data Analytics covers an important niche in the social network analytics field. This edited volume, contributed by prominent researchers in this field, presents a wide selection of topics on social network data mining such as Structural Properties of Social Networks, Algorithms for Structural Discovery of Social Networks and Content Analysis in Social Networks. This book is also unique in focussing on the data analytical aspects of social networks in the internet scenario, rather than the traditional sociology-driven emphasis prevalent in the existing books, which do not focus on the unique data-intensive characteristics of online social networks. Emphasis is placed on simplifying the content so that students and practitioners benefit from this book. This book targets advanced level students and researchers concentrating on computer science as a secondary text or reference book. Data mining, database, information security, electronic commerce and machine learning professionals will find this book a valuable asset, as well as primary associations such as ACM, IEEE and Management Science. Table of Contents Cover Social Network Data Analytics ISBN 9781441984616 Contents Preface AN INTRODUCTION TO SOCIAL NETWORK DATA ANALYTICS 1. Introduction 2. Online Social Networks: Research Issues 3. Research Topics in Social Networks 4. Conclusions and Future Directions STATISTICAL PROPERTIES OF SOCIAL NETWORKS 1. Preliminaries 1.1 Definitions 1.2 Data description 2. Static Properties 2.1 Static Unweighted Graphs 2.2 Static Weighted Graphs 3. Dynamic Properties 3.1 Dynamic Unweighted Graphs 3.2 Dynamic Weighted Graphs 4. Conclusion RANDOM WALKS IN SOCIAL NETWORKS AND THEIR APPLICATIONS: A SURVEY 1. Introduction 2. Random Walks on Graphs: Background 2.1 Random Walk based Proximity Measures 2.2 Other Graph-based Proximity Measures 2.3 Graph-theoretic Measures for Semi-supervised Learning 2.4 Clustering with random walk based measures 3. Related Work: Algorithms 3.1 Algorithms for Hitting and Commute Times 3.2 Algorithms for Computing Personalized Pagerank and Simrank 3.3 Algorithms for Computing Harmonic Functions 4. Related Work: Applications 4.1 Application in Computer Vision 4.2 Text Analysis 4.3 Collaborative Filtering 4.4 Combating Webspam 5. Related Work: Evaluation and datasets 5.1 Evaluation: Link Prediction 5.2 Publicly Available Data Sources 6. Conclusion and Future Work COMMUNITY DISCOVERY IN SOCIAL NETWORKS: APPLICATIONS, METHODS AND EMERGING TRENDS 1. Introduction 2. Communities in Context 3. Core Methods 3.1 Quality Functions 3.2 The Kernighan-Lin(KL) algorithm 3.3 Agglomerative/Divisive Algorithms 3.4 Spectral Algorithms 3.5 Multi-level Graph Partitioning 3.6 Markov Clustering 3.7 Other Approaches 4. Emerging Fields and Problems 4.1 Community Discovery in Dynamic Networks 4.2 Community Discovery in Heterogeneous Networks 4.3 Community Discovery in Directed Networks 4.4 Coupling Content and Relationship Information for Community Discovery 5. Crosscutting Issues and Concluding Remarks NODE CLASSIFICATION IN SOCIAL NETWORKS 1. Introduction 2. Problem Formulation 2.1 Representing data as a graph 2.2 The Node Classification Problem 3. Methods using Local Classifiers 3.1 Iterative Classification Method 4. Random Walk based Methods 4.1 Label Propagation 4.2 Graph Regularization 4.3 Adsorption 5. Applying Node Classification to Large Social Networks 5.1 Basic Approaches 5.2 Second-order Methods 5.3 Implementation within Map-Reduce 6. Related approaches 6.1 Inference using Graphical Models 6.2 Metric labeling 6.3 Spectral Partitioning 6.4 Graph Clustering 7. Variations on Node Classification 7.1 Dissimilarity in Labels 7.2 Edge Labeling 7.3 Label Summarization 8. Concluding Remarks 8.1 Future Directions and Challenges 8.2 Further Reading EVOLUTION IN SOCIAL NETWORKS: A SURVEY 1. Introduction 2. Framework 2.1 Modeling a Network across the Time Axis 2.2 Evolution across Four Dimensions 3. Challenges of Social Network Streams 4. Incremental Mining for Community Tracing 5. Tracing Smoothly Evolving Communities 5.1 Temporal Smoothness for Clusters 5.2 Dynamic Probabilistic Models 6. Laws of Evolution in Social Networks 7. Conclusion A SURVEY OF MODELS AND ALGORITHMS FOR SOCIAL INFLUENCE ANALYSIS 1. Introduction 2. Influence Related Statistics 2.1 Edge Measures 2.2 Node Measures 3. Social Similarity and Influence 3.1 Homophily 3.2 Existential Test for Social Influence 3.3 Influence and Actions 3.4 Influence and Interaction 4. Influence Maximization in Viral Marketing 4.1 Influence Maximization 4.2 Other Applications 5. Conclusion A SURVEY OF ALGORITHMS AND SYSTEMS FOR EXPERT LOCATION IN SOCIAL NETWORKS 1. Introduction 2. Definitions and Notation 3. Expert Location without Graph Constraints 3.1 Language Models for Document Information Retrieval 3.2 Language Models for Expert Location 3.3 Further Reading 4. Expert Location with Score Propagation 4.1 The PageRank Algorithm 4.2 HITS Algorithm 4.3 Expert Score Propagation 4.4 Further Reading 5. Expert Team Formation 5.1 Metrics 5.2 Forming Teams of Experts 5.3 Further Reading 6. Other Related Approaches 6.1 Agent-based Approach 6.2 Influence Maximization 7. Expert Location Systems 8. Conclusions A SURVEY OF LINK PREDICTION IN SOCIAL NETWORKS 1. Introduction 2. Background 3. Feature based Link Prediction 3.1 Feature Set Construction 3.2 Classification Models 4. Bayesian Probabilistic Models 4.1 Link Prediction by Local Probabilistic Models 4.2 Network Evolution based Probabilistic Model 4.3 Hierarchical Probabilistic Model 5. Probabilistic Relational Models 5.1 Relational Bayesian Network 5.2 Relational Markov Network 6. Linear Algebraic Methods 7. Recent development and Future Works PRIVACY IN SOCIAL NETWORKS: A SURVEY 1. Introduction 2. Privacy breaches in social networks 2.1 Identity disclosure 2.2 Attribute disclosure 2.3 Social link disclosure 2.4 Affiliation link disclosure 3. Privacy definitions for publishing data 3.1 k-anonymity 3.2 l-diversity and t-closeness 3.3 Differential privacy 4. Privacy-preserving mechanisms 4.1 Privacy mechanisms for social networks 4.2 Privacy mechanisms for affiliation networks 4.3 Privacy mechanisms for social and affiliation networks 5. Related literature 6. Conclusion VISUALIZING SOCIAL NETWORKS 1. Introduction 2. A Taxonomy of Visualizations 2.1 Structural Visualization 2.2 Semantic and Temporal Visualization 2.3 Statistical Visualization 3. The Convergence of Visualization, Interaction and Analytics 3.1 Structural and Semantic Filtering with Ontologies 3.2 Centrality-based Visual Discovery and Exploration 4. Summary DATA MINING IN SOCIAL MEDIA 1. Introduction 2. Data Mining in a Nutshell 3. Social Media 4. Motivations for Data Mining in Social Media 5. Data Mining Methods for Social Media 5.1 Data Representation 5.2 Data Mining A Process 5.3 Social Networking Sites: Illustrative Examples 5.4 The Blogosphere: Illustrative Examples 6. Related Efforts 6.1 Ethnography and Netnography 6.2 Event Maps 7. Conclusions TEXT MINING IN SOCIAL NETWORKS 1. Introduction 2. Keyword Search 2.1 Query Semantics and Answer Ranking 2.2 Keyword search over XML and relational data 2.3 Keyword search over graph data 3. Classification Algorithms 4. Clustering Algorithms 5. Transfer Learning in Heterogeneous Networks 6. Conclusions and Summary INTEGRATING SENSORS AND SOCIAL NETWORKS 1. Introduction 2. Sensors and Social Networks: Technological Enablers 3. Dynamic Modeling of Social Networks 4. System Design and Architectural Challenges 4.1 Privacy-preserving data collection 4.2 Generalized Model Construction 4.3 Real-time Decision Services 4.4 Recruitment Issues 4.5 Other Architectural Challenges 5. Database Representation: Issues and Challenges 6. Privacy Issues 7. Sensors and Social Networks: Applications 7.1 The Google Latitude Application 7.2 The Citysense and Macrosense Applications 7.3 Green GPS 7.4 Microsoft SensorMap 7.5 Animal and Object Tracking Applications 7.6 Participatory Sensing for Real-Time Services 8. Future Challenges and Research Directions MULTIMEDIA INFORMATION NETWORKS IN SOCIAL MEDIA 1. Introduction 2. Links from Semantics: Ontology-based Learning 3. Links from Community Media 3.1 Retrieval Systems for Community Media 3.2 Recommendation Systems for Community Media 4. Network of Personal Photo Albums 4.1 Actor-Centric Nature of Personal Collections 4.2 Quality Issues in Personal Collections 4.3 Time and Location Themes in Personal Collections 4.4 Content Overlap in Personal Collections 5. Network of Geographical Information 5.1 Semantic Annotation 5.2 Geographical Estimation 5.3 Other Applications 6. Inference Methods 6.1 Discriminative vs. Generative Models 6.2 Graph-based Inference: Ranking, Clustering and Semi-supervised Learning 6.3 Online Learning 7. Discussion of Data Sets and Industrial Systems 8. Discussion of Future Directions 8.1 Content-based Recommendation and Advertisements 8.2 Multimedia Information Networks via Cloud Computing AN OVERVIEW OF SOCIAL TAGGING AND APPLICATIONS 1. Introduction 1.1 Problems with Metadata Generation and Fixed Taxonomies 1.2 Folksonomies as a Solution 1.3 Outline 2. Tags: Why and What? 2.1 Different User Tagging Motivations 2.2 Kinds of Tags 2.3 Categorizers Versus Describers 2.4 Linguistic Classification of Tags 2.5 Game-based Tagging 3. Tag Generation Models 3.1 Polya Urn Generation Model 3.2 Language Model 3.3 Other Influence Factors 4. Tagging System Design 5. Tag analysis 5.1 Tagging Distributions 5.2 Identifying Tag Semantics 5.3 Tags Versus Keywords 6. Visualization of Tags 6.1 Tag Clouds for Browsing/Search 6.2 Tag Selection for Tag Clouds 6.3 Tag Hierarchy Generation 6.4 Tag Clouds Display Format 6.5 Tag Evolution Visualization 6.6 Popular Tag Cloud Demos 7. Tag Recommendations 7.1 Using Tag Quality 7.2 Using Tag Co-occurrences 7.3 Using Mutual Information between Words, Documents and Tags 7.4 Using Object Features 8. Applications of Tags 8.1 Indexing 8.2 Search 8.3 Taxonomy Generation 8.4 Public Library Cataloging 8.5 Clustering and Classification 8.6 Social Interesting Discovery 8.7 Enhanced Browsing 9. Integration 9.1 Integration using Tag Co-occurrence Analysis and Clustering 9.2 TAGMAS: Federated Tagging System 9.3 Correlating User Proles from Different Folksonomies 10. Tagging problems 10.1 Spamming 10.2 Canonicalization and Ambiguities 10.3 Other Problems 11. Conclusion and Future Directions 11.1 Analysis 11.2 Improving System Design 11.3 Personalized Tag Recommendations 11.4 More Applications 11.5 Dealing With Problems Index