|
Home Visualizations About Publications Contact Us Download |
|
Project Themes: Complex Social Systems Research Design and Methodology The Reality Mining Dataset Sociology in the 21st Century User Behavior Modeling and Prediction Relationship Inference Social Serendipity Organizational Dynamics Epidemiology and Information Dissemination Eigenbehaviors |
Complex Social SystemsDespite having a history of social science that spans several decades, we still have a relatively poor understanding of why humans behave the way they do. Given this ignorance, attempting to understand the complex collective behavior of groups of these idiosyncratic individuals, who make up organizations and even broader society, certainly seems even more daunting. Physicists have recently been quick to jump on the problem with their own set of tools, applying techniques such as statistical mechanics to ignore the micro-behavior of a system (ie: the speed of each individual particle in a balloon or individual in society), and rather provide guidelines for the behavior of the aggregate (ie: the air pressure in the balloon or the current cultural fad). Even in the early 70's, physicists began successfully mapping human movement in groups to Maxwell-Boltzmann kinetic theory of particle movement in gases [Henderson (1971)]. Today's physicists are now taking on much larger social phenomena: decision-making, contagion dissemination, the formation of alliances and organizations, as well as a wide range of other collective behavior [Newman (2001), Adamic & Huberman, (2003), Richardson & Domingos (2002), Albert & Barabasi (2002), Watts & Strogatz (1998), Eubank et al. (2004)].By continually logging and time-stamping information about a user's activity, location, and proximity to other users, the dynamics of large-scale human behavior can be measured. When deployed within a group of people working closely together, the phone and proximity log can be used to provide insight into the underlying dynamics of the organization. Furthermore, a dataset providing the proximity patterns and relationships within large groups of people has implications within the computational epidemiology communities, and may help build more accurate models of airborne pathogen dissemination, such as SARS, or other more innocuous contagions, such as simply the flow of gossip. It will also be striking to see the dynamics of the incoming students' social networks as they evolve over the course of their first semester.
© 2008 Massachusetts Institute of Technology
|