Network Evolution

Project Progress 

The team has continued its study of open source software communities using SourceForge, the largest online dataset containing over 250,000 individuals and 70,000 projects since 2006, from which we extracted dynamic information on the network affiliations of open source programmers and project teams.

The research focused on four areas, the first is on the impact of individual’s social capital on anticipating successful projects. Our analysis is capturing an individual’s network position before joining a project and after it is developed, with initial results indicating that those with high levels of brokerage or closure are able to anticipate projects that will become successful. We also examine the moderating effect of the environment on the individual’s network position. Work in this area is ongoing. The paper is being readied for submission.

Our second sub-project considered the influence of networks on the movement of programmers between projects, studying the creation and decay of collaborative ties. Although fluid ties are a key feature of distributed innovation communities it is not yet clear what drives movement between projects and what effect network resources play in these changes. A working paper is currently in development.

The third research project looks at the effects of social capital on project success. We are using the literature on entrepreneurial teams to examine the role of experience and network status both on venture success and the creation of effective teams. Although in its formative stages, we anticipate this work will result in recommendations for team building strategies.

The fourth research project examines how in the initial stages of project development, special efforts put into presenting and advertising a project influence the ability to recruit a team.

We are currently investigating possible industry partners and have already developed links with Trampoline Systems, a social network data mining company that helps companies assess the effectiveness of internal social networks. Trampoline Systems developed the Tech City map and we are currently exploring ways to work together.
Going Forward: In the next few months, we will continue our detailed analysis of these open innovation communities using Sourceforge data, focused on the how individuals accrue their social capital by anticipating successful projects in their early stages. We also plan to study migration patterns between different projects over time and to identify, extract and map the networks of members of these on-line communities over time and across project teams

Using this work we will prepare and document case histories of specific on-line communities, focusing on communities that span public and private sectors and their governance structure and evolution. This will allow us to prepare academic papers and expand our knowledge of network evolution in distributed innovation communities.