Masterarbeit
Developing a Tool for Automating Reproducibility Assessments for Repositories
Completion
2025/03
Research Area
Students
Jungsan Kim
student
Advisers
Dr. Sheeba Samuel
senior researcher
Prof. Dr.-Ing. Martin Gaedke
professor
Description
Reproducibility is a fundamental principle of scientific research, but there are still many challenges in reproducing published research results. In this study, we developed Re- proCheck, a system that automatically evaluates the reproducibility metrics of GitHub repositories. The tool extracts metadata and evaluates reproducibility metrics based on four main metrics: code quality, dependency management, data accessibility, and docu- mentation, each composed of detailed sub-metrics. The analysis results are visualized in a dashboard, which can be exported to RDF format and compared to averages across research areas to provide suggestions for improvement.
