If you’re an archaeologist interested in social networks in the past, what better way to keep on top of things than to see what’s going on in the present?

The Collective Dynamics Group at Columbia have a raft of papers online concerning their research – check them out.

The Group’s mandate is:

is the application of modern mathematical and computational techniques to problems relevant to the social sciences. Examples of current projects include the structure and evolution of social networks, the dynamics of disease epidemics and cultural fads, the role of social information in financial markets, and the use of the Internet as a tool for social science research. The group consists of graduate students and post-doctoral researchers from mathematics, sociology, and economic

Some of their major projects:

Interpersonal Influence, Contagion, and Collective Decision Making

People constantly influence each other in all facets of life. Social contagion is the spreading of ideas, rumors, and behavior through a population via interpersonal influences. Collective decisions are generated by a social contagion process which is (often greatly) augmented by the machinery of mass media. Consequently, understanding interpersonal influence is crucial to understanding the behavior of both individuals and groups. Our projects on influence are divided between online experiments and conceptual mathematical models. We have developed a generalized model of contagion that reconciles and extends previously disparate models of contagion from the social and biological sciences. We are interested in standard biological contagion alone since the collective behavior of people is almost always important in how diseases spread. For example, motivated by observations of the SARS outbreak in 2002, we are exploring the effect on a contagion’s spread due to people moving between subpopulations with some frequency. We are also currently developing an online experiment which will explore interpersonal influence in `cultural markets’ (markets for cultural products, such a books, music, celebrity, etc) and how individual behaviors aggregate to produce collective outcomes.

related papers

Social Search, Collective Problem Solving, and Organizational Robustness

The ability to solve problems collectively is central to the long term stability of any group of people, from a small business marketing a new product to nations confronting global economic crises. Real world collective problem solving is inherently a decentralized, distributed activity. When faced with a novel, ambiguous problem defined at the group level, individuals must determine how to coordinate their actions with others by exchanging ideas, knowledge, and questions. A key aspect of this coordination is search. How do invididuals find others who can at least partially answer or rephrase poorly specified problems? We approach this issue of what we call social search by building conceptual models and online experiments. For example, we have constructed a simple, sociologically plausible model of social networks that shows them to be searchable under general conditions. This is the so-called Small World hypothesis, the notion that two random individuals can find a way to connect to each other through a small number of intermediary contacts. For the past few years, we have been running a global small world experiment, where people send email to friends and acquaintances trying to find a sequence of contacts leading to `target’ individuals. In related work, we model modern organizations as reinforced hierarchical networks of individuals searching for information bearers among their peers. Being effective at collective problem solving leads to a tradeoff between specialized efficiency and flexible robustness.

related papers

Structure and Evolution of Social Networks

We explore social networks through data acquisition, theoretical model building, and online experiments. In the past few years, there has been a tremendous growth in the study of networks-at-large. While much is now understood about technological and physical networks, less is known about large-scale social networks. And for good reason: accurately determining who knows whom and to what degree is a difficult and highly time consuming task. However, with the advent of digital communication, we are now able to record vast numbers of interactions between individuals in large populations. In principle, certain channels of communication, such as e-mail exchange or instant messaging, can be recorded completely. Our current data acquisition projects focus specifically on collecting real-time e-mail interactions in combination with demographic data for large institutions. At the same time, we are working to build simple models of dynamic social networks, guided and informed by real data and online experiments. We also construct abstract models of evolving social networks to study basic concepts such as the emergence of cooperation. Our work is computer intensive and we have developed a suite of network analysis routines that will eventually be made publically available.

related papers