Deducing Case Ids from Unlabeled Event Log

Hey Guys,

I am interested in process discovery of unlabeled  event logs (there is no end-to-end processes, e.g., production processes).
Do you know any solution for such pre-processing task to convert an unlabeled to a labeled event log?
I am thinking of using unsupervised learning (clustering) based on other features such as input and output data?
Is there any support for such analysis in Prom? or any solution?

Thank you