Preface
Contents
Author Biographies
Part I Theory: Dilemmas of Model Building and Interpretation
Modeling the Complex Network of Social Interactions
1 Introduction
2 Characterization of Large-Scale Social Networks
2.1 Degree Distribution
2.2 Assortative Mixing by Degree
2.3 Clustering
2.4 Granovetterian Community Structure
2.5 Tie Formation and Fading
2.6 Multiplex Structure and Overlapping Communities
3 Modeling Granovetterian Community Structure
3.1 Weighted Social Network
3.2 Link Expiration
3.3 Multiplex Structure
4 Homophily and Structural Changes
5 Summary and Outlook
References
Formal Design Methods and the Relation Between Simulation Models and Theory: A Philosophy of Science Point of View
1 Introduction
2 Formal Design Methods for Agent-Based Simulation
2.1 An ODD Protocol for Abelson's and Bernstein's Early Work
2.2 Other Approaches
3 The `Non-statement View' of Structuralism
4 Translation into a New Simulation Model
4.1 Designing Agent Types, Relations and Functions
4.2 Results
5 Conclusion and Outlook
References
Part II Methodological Toolsets
The Potential of Automated Text Analytics in Social KnowledgeBuilding
1 Introduction
2 Challenges
3 A New Methodological Basis of Sociology
3.1 New Data Sources
3.2 A Brief Overview of NLP Methods
3.2.1 Pre-processing
3.2.2 Bag of Words and Beyond
3.3 The Goal of the Analysis and the Corresponding NLP Methods
3.3.1 Supervised Methods
3.3.2 Unsupervised Methods
3.3.3 Which Method to Choose
4 New Possibilities for Sociological Research
4.1 How to Approach Automated Text Analysis as a Social Scientist
4.2 Combining the New with the Traditional: Mixed Approaches
4.3 What the Approach Can Offer to Classic Sociological Questions
5 Summary
References
Combining Scientific and Non-scientific Surveys to Improve Estimation and Reduce Costs
1 Introduction
2 Why Bayesian?
3 Methodology and Modeling Approach
4 Application to the German Internet Panel
4.1 Probability and Nonprobability Data and Target Variables
4.2 Model Evaluation
4.3 Assessing Model Efficiency
5 Results
5.1 Variability in Mean Estimates of Height and Weight
5.2 Bias in Mean Estimates of Height and Weight
5.3 Mean-Squared Error (MSE) in Mean Estimates of Height and Weight
5.4 Efficiency Ratios for MSE and Variance
5.5 Potential Cost Savings for a Fixed MSE
6 Discussion
References
Harnessing the Power of Data Science to Grasp Insights About Human Behaviour, Thinking, and Feeling from Social Media Images
1 Introduction
2 Social Media Images: Pathways Between Traditional Social Science and Computational Social Science
2.1 The Relationship Between Social Media Images and Personality, Depression, Emotions, and Other Psychological Constructs: A Literature Review
2.1.1 Search Strategy, Eligibility Criteria, and Results
2.1.2 Personality and Individual Differences
2.1.3 Depression and Mental Health
3 An Intuitive Introduction to Image Processing and Analysis
3.1 Pixel-Level Features
3.2 Face Detection
3.3 Convolutional Neural Networks (CNNs)
4 Conclusions
References
Validating Simulation Models: The Case of Opinion Dynamics
1 Introduction
2 Types of Validity: Validation Against Empirical Data and Against Stylised Facts
2.1 Types of Validity
3 Models of Opinion and Attitude Dynamics
3.1 One-Dimensional Models
3.2 Multiple Attitude Dynamics
3.3 A Two-Dimensional Model Along the Lines of Hegselmann-Krause and Deffuant-Weisbuch
4 Opinion and Attitude Dynamics: Empirical Findings
4.1 Reported Concerns in the German Socio-Economic Panel from 1984 till 2016
4.2 Party Scalometers in the German Election Panel 2016β2018
4.3 A First Conclusion from the Empirical Evidence
5 Calibrating and Validating the Model Against the Empirical Cases
5.1 The GLES Version of the Model
5.2 The GSOEP Version of the Model
5.3 The Original Versions of the Model
5.3.1 Initialisation and Other Stochastic Effects with a Normal Distribution
5.3.2 Initialisation and Other Stochastic Effects with a Uniform Distribution
6 Conclusion
Appendix: Results of Data Transformations
GSOEP Variables About Concerns
German Longitudinal Election Study (GLES): Scalometers and Party Preferences of the Campaign Panel 2017
Politbarometer: Selected Results from Scalometers from 1994 to 2016
References
Part III New Look on Old Issues: Research Domains Revisited by Computational Social Science
A Spatio-Temporal Approach to Latent Variables: Modelling Gender (im)balance in the Big Data Era
1 Introduction
2 The Gender Data Revolution: Setting a New Frontier in Engendered Statistics
3 The Rise of Computational Approaches from Recent Statistical Advances
3.1 The Multivariate Latent Markov Model for Spatio-Temporal Studies at a Glance
4 Towards a Computational Approach to the Gender Gap Issue in the Network Age
4.1 When Current Gender Gap Indexes Do Not Support Disambiguation of Societal Trends
5 Conclusions
Appendix
References
Agent-Based Organizational Ecologies: A Generative Approach to Market Evolution
1 Introduction
2 Industrial Organization and Computation
3 Organizational Ecology and Simulation Modeling
4 NK Models and Industry Dynamics
5 Generative Processes of Markets
6 Explanation and Generative Processes
7 Conclusions
References
Networks of the Political Elite and Political Agenda Topics: Creation and Analysis of Historical Corpora Using NLP and SNA Methods
1 Introduction
2 Data and Analysis
2.1 Textual Sources and Analysis Approaches
2.2 Text Network Analysis (TNA)
2.3 Classification Methods
2.4 Ontology-Based Classification
2.5 Classification Using an AI Classifier
2.6 The Resulting Relational Network
2.7 A New Approach to the Political Elite of This Era
3 Conclusions
References
Participatory Budgeting: Models and Approaches
1 Introduction
1.1 Outline
2 Mathematical Formulation
2.1 Decision Space and Popular PB Models
2.1.1 Bounded Discrete PB (Combinatorial PB)
2.1.2 Discrete PB
2.1.3 Divisible PB
2.1.4 Unbounded Divisible PB (Portioning)
2.2 Preference Modeling and Ballot Design
2.2.1 Preference Modeling
2.2.2 Ballot Design
2.3 Vote Aggregation
2.3.1 Welfare Maximization
2.3.2 The Axiomatic Approach
2.3.3 Fairness
3 Discrete Participatory Budgeting
3.1 Review of the Literature on Settings Related to Discrete PB
3.2 Approaches to Discrete PB
3.2.1 Welfare Maximization
3.2.2 Elicitation
3.2.3 Incentives
3.2.4 Axiomatic Desiderata
3.2.5 Fairness
3.2.6 Other Approaches
4 Divisible Participatory Budgeting
4.1 Review of the Literature on Settings Related to Divisible PB
4.2 Approaches to Divisible PB
4.2.1 Welfare Maximization
4.2.2 Fairness
4.2.3 Incentives
5 Extensions and Future Directions
References
From Durkheim to Machine Learning: Finding the Relevant Sociological Content in Depression and Suicide-Related Social Media Discourses
1 Introduction
2 Recent Studies of the Field
3 Data and Methods
4 Ways to Find and Analyze the Relevant Content
4.1 The Application of Topic Model
4.2 The Application of Word-Embedding
5 Discussion
Appendix
References
Epilogue
Changing Understanding in Algorithmic Societies: Exploring a New Social Reality with the Tools of Computational Social Science
1 Algorithmic Societies
2 The Changing Perception of Reality
3 Algorithmic Awareness and Our Changing View on Privacy
4 New Possibilities and New Barriers of Information Access
5 The Social Installation of Algorithmic Entities
6 The Changing Role of Social Science
References
Index