Publication Abstract

Pickard, G., Khazan, R., Fuller, B., Cooley, J., A Corpus of Group Dynamics Data from Internet Chatrooms, 2010 International Conference on Network Science, May 12-14, 2010.


What follows is a presentation of a methodology for extracting useful group dynamic data from the logs of group-chat sessions, and also the corpus of data that results from applying this methodology to a large, publicly-available collection of Internet chatlogs. This work is motivated by the lack of a large, consistent data-set in the public domain that might be used to apply quantitative methods to the study of group dynamics. While there has been a wealth of research regarding social networks, much of which is precise enough to allow for quantitative measurement, virtually none of it deals with the dynamics of groups in flux, and there does not exist a common data set, or even a widely accepted data format, that could be used to support such research.

Our research focuses on group oriented communications, and in particular on the problem of securing communication among dynamic groups. We present this data-set as an example of the kind of data which we have found useful in the evaluation and improvement of our algorithms. It is our hope that by making the data available to the public, we will encourage further research in this area – and that by making the methodology for producing such data available, we will encourage others to collect and share similar data.


top of page