There has been intensifying interest over the past decade in analyzing cyberspace to detect the values, beliefs, and social relationships of the people who use it and generate its content. However, despite the intrinsically social nature of this phenomenon, much more can be done to integrate the social and cultural theories developed by social scientists with the technical approaches developed by computer scientists. This project will result in a set of analytical methods and software based on social science social network and cultural theories, melded with computer science data mining techniques, which will provide ways in which to analyze virtual communities with as much, or greater, ease as we can analyze the “real world” social communities that they represent. This in turn will provide data and methods to predict the ideas, values, and actions of social communities, as well as the ways in which different communities interact with one another. More specifically, it will build upon prior technologies: a specialized theory-centered crawler being developed at the University of Hawaii, and advanced data mining techniques for identifying the “long tail” of virtual communities from Arizona State University, but will integrate them and incorporate a much wider range of social science models, thus producing a general tool that can significantly help a wide range of researchers studying virtual communities.
Funded by: UH/AFOSR (FA9550-09-1-0261), 02/15/2009 - 11/30/2011.
Arizona State University, Data Mining and Machine Learning Lab (DMML),
Last Update: Sept. 10, 2009