Answers are grounded in this paper and cite their sources.
Issue trackers record every stage an issue passes through, but these histories are irregular and incomplete, which makes it hard to know what process discovery can deliver for software analytics. We convert the JIRA histories of 175 JBoss/Red Hat sub-projects into an event log of 17,671 status events and 5,759 issues. We then apply the Alpha Miner, the Heuristics Miner, the Inductive Miner and a frequency-weighted Heuristics Net, and we replace activities with developers to build a handover-of-work network. A qualitative comparison shows a strong create-resolve-close path, a long tail of rare activities and recurring rework, and only the Alpha Miner fails to connect the reopening loop. The handover network forms separate clusters, its links run mostly through comments, and it is concentrated around a few developers. Choices made while building the event log shape what is discovered, so the log itself is a modelling decision. Quantitative quality measures remain future work.
No takes yet. Share an insight, caveat, or question.
Subedi et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: