SOCIAL NETWORK DATA ANALYTICS

Transcription

1 SOCIAL NETWORK DATA ANALYTICS

2

3 SOCIAL NETWORK DATA ANALYTICS Edited by CHARU C. AGGARWAL IBM T. J. Watson Research Center, Yorktown Heights, NY 10598, USA Kluwer Academic Publishers Boston/Dordrecht/London

4 Contents Preface 1 An Introduction to Social Network Data Analytics 1 Charu C. Aggarwal 1. Introduction 1 2. Online Social Networks: Research Issues 5 3. Research Topics in Social Networks 8 4. Conclusions and Future Directions 13 References 14 2 Statistical Properties of Social Networks 17 Mary McGlohon, Leman Akoglu and Christos Faloutsos 1. Preliminaries Definitions Data description Static Properties Static Unweighted Graphs Static Weighted Graphs Dynamic Properties Dynamic Unweighted Graphs Dynamic Weighted Graphs Conclusion 39 References 40 3 Random Walks in Social Networks and their Applications: A Survey 43 Purnamrita Sarkar and Andrew W. Moore 1. Introduction Random Walks on Graphs: Background Random Walk based Proximity Measures Other Graph-based Proximity Measures Graph-theoretic Measures for Semi-supervised Learning Clustering with random walk based measures Related Work: Algorithms Algorithms for Hitting and Commute Times Algorithms for Computing Personalized Pagerank and Simrank 60 xiii

5 vi SOCIAL NETWORK DATA ANALYTICS 3.3 Algorithms for Computing Harmonic Functions Related Work: Applications Application in Computer Vision Text Analysis Collaborative Filtering Combating Webspam Related Work: Evaluation and datasets Evaluation: Link Prediction Publicly Available Data Sources Conclusion and Future Work 69 References 71 4 Community Discovery in Social 79 Networks: Applications, Methods and Emerging Trends S. Parthasarathy, Y. Ruan and V. Satuluri 1. Introduction Communities in Context Core Methods Quality Functions The Kernighan-Lin(KL) algorithm Agglomerative/Divisive Algorithms Spectral Algorithms Multi-level Graph Partitioning Markov Clustering Other Approaches Emerging Fields and Problems Community Discovery in Dynamic Networks Community Discovery in Heterogeneous Networks Community Discovery in Directed Networks Coupling Content and Relationship Information for Community Discovery Crosscutting Issues and Concluding Remarks 102 References Node Classification in Social Networks 115 Smriti Bhagat, Graham Cormode and S. Muthukrishnan 1. Introduction Problem Formulation Representing data as a graph The Node Classification Problem Methods using Local Classifiers Iterative Classification Method Random Walk based Methods Label Propagation Graph Regularization Adsorption Applying Node Classification to Large Social Networks Basic Approaches Second-order Methods Implementation within Map-Reduce 138

6 Contents vii 6. Related approaches Inference using Graphical Models Metric labeling Spectral Partitioning Graph Clustering Variations on Node Classification Dissimilarity in Labels Edge Labeling Label Summarization Concluding Remarks Future Directions and Challenges Further Reading 146 References Evolution in Social Networks: A Survey 149 Myra Spiliopoulou 1. Introduction Framework Modeling a Network across the Time Axis Evolution across Four Dimensions Challenges of Social Network Streams Incremental Mining for Community Tracing Tracing Smoothly Evolving Communities Temporal Smoothness for Clusters Dynamic Probabilistic Models Laws of Evolution in Social Networks Conclusion 169 References A Survey of Models and Algorithms for Social Influence Analysis 177 Jimeng Sun and Jie Tang 1. Introduction Influence Related Statistics Edge Measures Node Measures Social Similarity and Influence Homophily Existential Test for Social Influence Influence and Actions Influence and Interaction Influence Maximization in Viral Marketing Influence Maximization Other Applications Conclusion 208 References A Survey of Algorithms and Systems for Expert Location in Social Networks 215 Theodoros Lappas, Kun Liu and Evimaria Terzi 1. Introduction 216

7 viii SOCIAL NETWORK DATA ANALYTICS 2. Definitions and Notation Expert Location without Graph Constraints Language Models for Document Information Retrieval Language Models for Expert Location Further Reading Expert Location with Score Propagation The PageRank Algorithm HITS Algorithm Expert Score Propagation Further Reading Expert Team Formation Metrics Forming Teams of Experts Further Reading Other Related Approaches Agent-based Approach Influence Maximization Expert Location Systems Conclusions 235 References A Survey of Link Prediction 243 in Social Networks Mohammad Al Hasan and Mohammed J. Zaki 1. Introduction Background Feature based Link Prediction Feature Set Construction Classification Models Bayesian Probabilistic Models Link Prediction by Local Probabilistic Models Network Evolution based Probabilistic Model Hierarchical Probabilistic Model Probabilistic Relational Models Relational Bayesian Network Relational Markov Network Linear Algebraic Methods Recent development and Future Works 269 References Privacy in Social Networks: A Survey 277 Elena Zheleva and Lise Getoor 1. Introduction Privacy breaches in social networks Identity disclosure Attribute disclosure Social link disclosure Affiliation link disclosure Privacy definitions for publishing data k-anonymity 288

8 Contents ix 3.2 l-diversity and t-closeness Differential privacy Privacy-preserving mechanisms Privacy mechanisms for social networks Privacy mechanismsfor affiliation networks Privacy mechanismsfor social and affiliation networks Related literature Conclusion 302 References Visualizing Social Networks 307 Carlos D. Correa and Kwan-Liu Ma 1. Introduction A Taxonomy of Visualizations Structural Visualization Semantic and Temporal Visualization Statistical Visualization The Convergence of Visualization, Interaction and Analytics Structural and Semantic Filtering with Ontologies Centrality-based Visual Discovery and Exploration Summary 322 References Data Mining in Social Media 327 Geoffrey Barbier and Huan Liu 1. Introduction Data Mining in a Nutshell Social Media Motivations for Data Mining in Social Media Data Mining Methods for Social Media Data Representation Data Mining - A Process Social Networking Sites: Illustrative Examples The Blogosphere: Illustrative Examples Related Efforts Ethnography and Netnography Event Maps Conclusions 345 References Text Mining in Social Networks 353 Charu C. Aggarwal and Haixun Wang 1. Introduction Keyword Search Query Semantics and Answer Ranking Keyword search over XML and relational data Keyword search over graph data Classification Algorithms 366

9 x SOCIAL NETWORK DATA ANALYTICS 4. Clustering Algorithms Transfer Learning in Heterogeneous Networks Conclusions and Summary 373 References Integrating Sensors and Social Networks 379 Charu C. Aggarwal and Tarek Abdelzaher 1. Introduction Sensors and Social Networks: Technological Enablers Dynamic Modeling of Social Networks System Design and Architectural Challenges Privacy-preserving data collection Generalized Model Construction Real-time Decision Services Recruitment Issues Other Architectural Challenges Database Representation: Issues and Challenges Privacy Issues Sensors and Social Networks: Applications The Google Latitude Application The Citysense and Macrosense Applications Green GPS Microsoft SensorMap Animal and Object Tracking Applications Participatory Sensing for Real-Time Services Future Challenges and Research Directions 407 References Multimedia Information Networks in Social Media 413 Liangliang Cao, GuoJun Qi, Shen-Fu Tsai, Min-Hsuan Tsai, Andrey Del Pozo, Thomas S. Huang, Xuemei Zhang and Suk Hwan Lim 1. Introduction Links from Semantics: Ontology-based Learning Links from Community Media Retrieval Systems for Community Media Recommendation Systems for Community Media Network of Personal Photo Albums Actor-Centric Nature of Personal Collections Quality Issues in Personal Collections Time and Location Themes in Personal Collections Content Overlap in Personal Collections Network of Geographical Information Semantic Annotation Geographical Estimation Other Applications Inference Methods Discriminative vs. Generative Models Graph-based Inference: Ranking, Clustering and Semi-supervised Learning 428

10 Contents xi 6.3 Online Learning Discussion of Data Sets and Industrial Systems Discussion of Future Directions Content-based Recommendation and Advertisements Multimedia Information Networks via Cloud Computing 434 References An Overview of Social Tagging and Applications 447 Manish Gupta, Rui Li, Zhijun Yin and Jiawei Han 1. Introduction Problems with Metadata Generation and Fixed Taxonomies Folksonomies as a Solution Outline Tags: Why and What? Different User Tagging Motivations Kinds of Tags Categorizers Versus Describers Linguistic Classification of Tags Game-based Tagging Tag Generation Models Polya Urn Generation Model Language Model Other Influence Factors Tagging System Design Tag analysis Tagging Distributions Identifying Tag Semantics Tags Versus Keywords Visualization of Tags Tag Clouds for Browsing/Search Tag Selection for Tag Clouds Tag Hierarchy Generation Tag Clouds Display Format Tag Evolution Visualization Popular Tag Cloud Demos Tag Recommendations Using Tag Quality Using Tag Co-occurrences Using Mutual Information between Words, Documents and Tags Using Object Features Applications of Tags Indexing Search Taxonomy Generation Public Library Cataloging Clustering and Classification Social Interesting Discovery Enhanced Browsing Integration 485