Research Team Including UMass Amherst’s Charlie Schweik and Brenda Bushouse Receives $3.4 Million in NSF Grants to Study Open-Source Software

Image

A research team including Charlie Schweik and Brenda Bushouse of the University of Massachusetts Amherst has been awarded $3.4 million in grants from the National Science Foundation’s Growing Convergence Research program for their project, “Jumpstarting Successful Open-Source Software Projects with Evidence-based Rules and Structures.”

Schweik, professor of public policy and environmental conservation, and Bushouse, associate professor of political science and public policy, will work in collaboration with Vladimir Filkov and Seth Frey of the University of California Davis to discern the socio-technical structural and governance conditions under which internet-based open-source software (OSS) projects are most—and least—effective, and provide actionable knowledge to OSS developers and to the nonprofit organizations that support them.

The NSF Growing Convergence Research program funds projects that address complex research problems, particularly ones focused on societal needs, by bringing together experts from across disciplines. The UMass/UC Davis project brings together Bushouse’s expertise in the governance of nonprofit organizations, Schweik’s in internet-based collaboration and open-source software, Filkov’s in software engineering and Frey’s in human-decision behavior in complex social environments. Together they will develop a framework for assessing the likelihood of success of OSS projects.

“Open-source software is a multi-billion-dollar industry,” Schweik said. “Over 80% of businesses, including all major tech companies, rely on OSS, and most people use it in their day-to-day digital lives—browsing the web, editing documents, banking, hosting websites—often without realizing it.

“While this popularity attracts many programmers to create open-source projects, more than 90% of OSS projects are eventually abandoned, especially smaller and start-up phase projects,” Schweik continued.

In their project, the UMass and UC Davis researchers will connect parallel streams of research about open-source software development from software engineering, data science and the social sciences to generate new data and knowledge to help understand whether OSS projects, especially nascent ones, are likely to be successful and self-sustaining. The team will focus on nonprofit-supported projects, such as those created under the Apache Software Foundation, and will engage open-source communities throughout the project, both to inform the research and to explore applications and results. The work will advance the theory and practice of software engineering, especially as it relates to understanding the success and effectiveness of OSS projects, goals that are critical to national competitiveness. The broader impact of the project is its potential to strengthen a technology that has become woven into the fabric of society.

“Some of my past research focused on the governance and management of OSS projects to understand what leads some projects to continue to be worked on, but the vast majority of them to become abandoned,” Schweik said. “This project builds upon that work but allows our team to build very large data sets of projects, thanks to the data science capabilities at UC Davis. Bushouse brings her expertise in nonprofit governance and qualitative methods to examine the much richer and more complex governance and management that occurs when OSS projects are part of incubator programs created by nonprofit organizations that establish the rules for participation.”

“The puzzle in this project is how the parent nonprofit organization’s membership rules affect the OSS collaboration process and the outcomes of those processes,” Bushouse said.

Through interviews and survey responses, Schweik and Bushouse will identify the multi-level rules, norms and strategies that ultimately determine whether an OSS project continues to be worked on, supported and further developed, or whether it is abandoned.