The Duke Network Analysis Center

The purpose of the Duke Network Analysis Center is to build a world-class network science & analysis program at Duke to: (a) help make visible the cutting-edge network scholarship currently ongoing on campus, (b) promote new collaborations in network science, (c) introduce new researchers to network science and train them in its methods and applications, (d) provide a research service in network analysis methods to the wider Duke community, and (e) enhance Duke’s position as a leader in the research triangle and throughout the nation in this exciting interdisciplinary field.

Why a network center at Duke?

Duke University is perfectly situated to capitalize on this exciting research frontier, and there is every reason to believe that Duke is among one of the nation’s leading sites for network research.  A center for network research will create linkages among the leading network researchas already active on campus to help each other do better work with greater visibility. 

The ultimate purpose of a Duke Network Analysis Center is to produce world-class network research that would be difficult to accomplish by investigators working alone.  We see three key activities of a network center: (1) Building a strong network research community to share and develop ideas, (2) helping facilitate new funded research projects by linking investigators with complementary skills, (3) providing training and technical analysis support for new network-related research projects.   

A key aspect of our community building is our weekly seminar series, which meets over lunch to discuss ongoing affiliate projects.  We are helping build new research by providing RA support for new projects and --  coming soon! – will be providing seed grants to help jump start new projects.  We provide regular mini-courses in network analysis (schedule TBA) and provide ad-hoc support to researchers seeking help with network projects.

Intellectual Background: A very brief history of network science.

There has been a quiet revolution across many areas of science recently; shifting our investigations to the patterns of connection linking system elements rather than the properties of such elements.  While scientific advance in the last 25 years rested on our capacity to isolate and divide, progress in the next 25 years likely rests on understanding how those elements are connected; how these connections shape outcomes; and how such multi-level complex systems evolve. This growing body of research on connected systems has touched most areas of research; literally from Anthropology to Zoology.  The figure at right, for example, traces the dramatic growth in the study of networks across the natural and social sciences and the humanities.  For a good overview of problems and approaches, see the special issue of Science published 24 July 2009 (Vol 325).  This increase in scientific activity has drawn the attention of funding agencies as well.  This year the NIH has announced a new program in the behavioral social sciences branch devoted to understanding fundamental elements of network science, the NSF has recently invested nearly $50M in their human social dynamics program and continues to invest through a strong network component to many of the NSF’s cross-cutting programs.


Social network approaches have figured prominently in sociology, psychology and anthropology since the early 1930s, when Moreno first introduced the “sociogram” as a tool for analyzing youth behavior (see figure at right).  He demonstrated that mapping the relations facilitated treatment for individual behavior problems and helped manage relational conflict in the group homes.   But most importantly, it signaled a new insight: that previously “invisible” elements of social life could be made explicit and studied directly.  In the ensuing eighty years, many social scientists have argued that the essence of social life turns on the emergent features created through such networks.  With spotty growth throughout the century, the field of social network analysis officially organized in the mid 1970s, and has now expanded throughout the social sciences.  Similar insights into the importance of connectivity across settings have sparked a rapid growth in the use of network tools in physics, computer science, biology, ecology and medicine.  The tools initially developed to map connections between people can also be used to map connections among proteins, diseases, servers or species.  Work on networks now appears regularly in Science, Nature, and PNAS.   Importantly, this field is a truly interdisciplinary science, where insights and tools developed in the natural sciences can be leveraged in the social sciences and vice versa, making collaboration and communication across such fields extremely important. 

Organization & Funding

The center is funded with collaborative funds from the Provost’s office, Dean’s office, SSRI, and investigator departments and is lead by James Moody (Sociology) with an ad hoc advisory board consisting of:  David Banks (Statistics), Phil Costanzo (Psychology), Jonathon Cummings (Fuqua), Jeffery Forbes (Computer Science), Rachel Kranton (Economics), Miller McPherson (Sociology), Peter Mucha (Math, UNC), Joshua Socolar  (Physics), Dean Urban (Nicholas School), and Michael Ward (Political Science).


Tag Cloud image