May I ask what algorithm is used to mine the model that is now excavated? I can already run the community version and import the project into eclipse to generate a process model, but I don’t know how to modify the source code to use my own developed process mining algorithm for modeling. Looking forward to your reply.
Apromore extracts a full DFG (directly-follows graph) from a log by computing the consecutive activity instance pairs and aggregating their performance metrics (e.g. case frequency, total frequency, etc.).
The user can then filter the DFG to the desired level (for this, arcs are dropped one by one while not dropping arcs that would disconnect the graph).
When you ask for the BPMN model, Apromore uses the split miner algorithm to turn the DFG into a BPMN model. The Split Miner algorithm is documented in this research paper:
The Apromore code base is not really designed to change the discovery algorithm. It would be a lot of effort to change it.