Security Anomaly Detection in Enterprise GitHub
A. Jordan, Yong Chen · 2024
Enterprises that build software on a large scale have complexities that are unique, where the software delivery community may scale to numbers that are difficult to oversee, either from a software engineering or security perspective. The ability to detect anomalies in Enterprise GitHub can open a new opportunity to gather insights on misuse, risky behaviors, and suspicious behaviors in software engineering services. An effective method to detect anomalies in a large enterprise is developed and a foundational step in maturing methods to improve software engineering practices and security at scale. This method utilizes machine learning and security indicators in audit logs to effectively identify anomalies at scale.