Organizational Influence on Security Development in Open-Source Software Projects

Abstract

Increasing technological complexity, intensified competition, and security requirements have driven open-source software (OSS) projects to become a crucial part of organizations' software development. This study focuses on the OSS project TensorFlow (TF) and uses a case study to examine how organizations and their associated developers collaborate to identify, fix and prevent security vulnerabilities. Social Network Analysis (SNA) of archived security data from software repositories is used to gain insight into security activities. The study examines the internal structure and evolution of security code collaboration, organizational networks, and top organizational contributors to TF. It also examines productivity, homophily, development diversity, and turnover rates among developers across various software releases. The in-depth insights from this research enhance our understanding of collaborative patterns in OSS communities within open software ecosystems, particularly in the security context.