A METHODOLOGY FOR CLUSTERING GITHUB PROJECTS BY STAGES OF THEIR DEVELOPMENT

This article presents a methodology for clustering GitHub projects based on their development stages using repository characteristics from a Kaggle dataset. By analyzing features such as star count (rating) and fork count (reuse), projects are classified into three types (prototype, tool, library) using the k-nearest neighbors (KNN) algorithm. The study includes an overview of GitHub’s theoretical foundations, justification for clustering effectiveness in project lifecycle analysis, a description of the developed algorithm, and implementation details in Python with specialized libraries. The results demonstrate the practical applicability of the method for optimizing open-source project support and trend analysis.