Revealing code change propagation channels by evolution history mining

Daihong Zhou, Yijian Wu, Xin Peng, Jiyue Zhang, Ziliang Li

Abstract: Highlights•We propose a new approach to capture relevant code changes in and across commits.•We propose a graph-based mining approach for finding Change Propagation Channels.•We discover 40 Frequent CPCs in 5 OSS projects, useful for change impact prediction.•We release a graph database of code entity relationships captured in 5 OSS projects.