social network analysis

social network analysis
The Opte Project · CC BY 2.5 · source
Name	Social network analysis
Discipline	Sociology; Anthropology; Computer science; Statistics
Subdisciplines	Network science; Graph theory; Computational social science
Notable persons	Jacob Moreno; Stanley Wasserman; Linton C. Freeman; Duncan J. Watts; Mark Granovetter

Contents

social network analysis

Social network analysis examines relational structures among actors using graph-based models and quantitative methods. It intersects with Sociology, Anthropology, Computer science, Statistics and draws on contributions from figures associated with Columbia University, Harvard University, University of Chicago, Stanford University and London School of Economics. Practitioners apply techniques developed in contexts such as the Manhattan Project-era studies, Cold War organizational research at RAND Corporation, and modern work at institutions like Google and Microsoft Research.

Overview

SNA represents actors (individuals, organizations, states) as nodes and their ties as edges within a graph theory framework linked to methods from Probability theory, Linear algebra, Matrix theory, Complex networks and Operations Research. Core goals include mapping structural positions relevant to phenomena studied at Princeton University, Yale University, Massachusetts Institute of Technology and University of Oxford. The field informs policy at agencies such as United Nations, World Bank, European Commission and corporations like Facebook and Amazon.

Early formal roots trace to work by Jacob Moreno and the Sociometry movement, later expanded by scholars at Harvard University and Columbia University during the mid-20th century. Influential theoretical contributions came from Mark Granovetter on weak ties, Ronald Burt on structural holes (affiliated with University of Chicago), and mathematical formalizations by Linton C. Freeman and Stanley Wasserman. Interactions with Paul Erdős and Alfréd Rényi informed random-graph models originating at Eötvös Loránd University and evolving through studies at Bell Labs and Los Alamos National Laboratory. Later cross-pollination with Duncan J. Watts and Albert-László Barabási at Cornell University and Northeastern University produced small-world and scale-free network theories influential in studies at Imperial College London and ETH Zurich.

Analytic methods include descriptive graph measures, stochastic models and algorithmic approaches developed in computer science contexts such as Carnegie Mellon University and University of California, Berkeley. Common metrics: degree centrality linked to work at Columbia University; betweenness centrality related to studies at MIT; closeness centrality used in analyses at Princeton University; eigenvector centrality developed in line with research at Bell Labs and AT&T Laboratories. Community detection algorithms draw on techniques tested at Google and Yahoo! Research and formal models such as exponential random graph models (ERGM) associated with Pennsylvania State University and Indiana University. Dynamic and temporal network models connect to projects at Los Alamos National Laboratory and Sandia National Laboratories.

Data sources span archival records from National Archives and Records Administration, ethnographic datasets collected in fieldwork by scholars at University of California, Los Angeles and digital traces harvested from platforms like Twitter, LinkedIn, Instagram, Reddit and Wikipedia (projects at Wikimedia Foundation). Preprocessing tasks use tools developed at Google and Microsoft Research and standards promoted by Institute of Electrical and Electronics Engineers and International Organization for Standardization bodies. Challenges include entity resolution addressed in collaborations with IBM Research and de-duplication methods from Siemens and Oracle.

SNA supports research and practice across domains: public health interventions studied at Centers for Disease Control and Prevention and World Health Organization; organizational analysis at McKinsey & Company and Boston Consulting Group; terrorism and security studies at RAND Corporation and U.S. Department of Defense; innovation diffusion analyses connected to Bell Labs and IBM; political network mapping by teams linked to Cambridge Analytica controversies and research at Harvard Kennedy School. It informs urban studies in projects at New York University and University College London, supply-chain analyses in firms like Walmart and Maersk, and ecological network research by groups at Smithsonian Institution and Scripps Institution of Oceanography.

Limitations include sampling biases noted in studies by American Sociological Association members, measurement error issues explored in work affiliated with National Science Foundation, and model misspecification discussed at conferences hosted by Association for Computing Machinery and Institute of Mathematical Statistics. Ethical considerations involve privacy debates litigated in cases before European Court of Human Rights and regulatory responses from Federal Trade Commission and European Commission; concerns over algorithmic unfairness have been raised in forums at Stanford University's centers and Harvard University's ethics initiatives. Responsible practice references guidelines from UNESCO and professional codes at American Statistical Association and Association for Computing Machinery.

Category:Network analysis