Uncovering the Hidden Patterns of Contributor Engagements in Active and Inactive GitHub Projects

In today's digital era, collaborative software development has become increasingly important, with GitHub serving as one of the primary platforms. However, not all projects on GitHub continue to thrive; many become inactive and ultimately turn into “dead projects.” This research aims to identify the patterns of contributor engagement and their support in shaping the GitHub project activity status. A qualitative approach was used to analyze these impacts, using topological data analysis (TDA) and K-means clustering. This study analyzed 96 OSS projects (70 active projects and 26 inactive ones) on GitHub, comprising contributors, participation, and popularity aspects. The results indicate that projects dominated by long-term contributors with high participation and popularity tend to be more stable and develop sustainably. In contrast, projects with a majority of short-term contributors with low engagement and popularity are more prone to inactivity. In addition, the clustering analysis grouped the projects into two clusters with distinct characteristics, where Cluster 0 represents active projects characterized by high contributions, significant popularity, and regular updates. Cluster 1 reflects projects with low or inactive activity levels, marked by minimal contributions, low popularity, and infrequent updates. This study provides comprehensive information on contributor patterns in determining the activity performance of open-source software projects.