Analysis of the Structure and Evolution of an Open-Source Community

Open-source processes are based on the paradigm of self-organized communities as opposed to the traditional hierarchical teams. These processes have not only been successful in the software development domain but are also increasingly being used in the development of physical products. In order to successfully adapt open-source processes to product realization, there is a need to understand how open-source communities self-organize and how this impacts the development of products. Toward the direction of fulfilling this need, we present an analysis of an existing open-source community involved in developing a web-based content-management platform, Drupal. The approach is based on the analysis of networks using techniques such as social network analysis, degree distribution, and hierarchical clustering. Openly available information on the Drupal web-site is utilized to perform the analysis of the community. The data are transformed into two weighted undirected networks: networks of people and networks of Drupal modules. Both the structures of these networks and their evolution during the past 6 years are studied. Based on the analysis, it is observed that the structure of the Drupal community has the characteristics of a scale-free network, which is similar to many other complex networks in diverse domains. Key trends in the evolution of the networks are identified. Finally, a predictive model is presented to provide potential explanations for the observed structures and evolutionary trends. [DOI: 10.1115/1.3615677]