US Patent 11301244 Machine learning based tracking of derivaitive code

In an approach for using machine learning to track programming code derivatives of source code, a processor captures the source code to track iterations of the source code. A processor detects a change of the source code. A processor analyzes derivative code from the source code for correlation with the source code based on similarity. A processor determines that one or more functions of the derivative code are related to the change of the source code based on the correlation. A processor highlights the related one or more functions of the derivative code.