Process mining is a set of analysis made on any enterprise data logs or events, in order to discover processes and compute some statistics or business analytic queries. Applied to Bonita, we use BPM event logs stored in Bonita Database to compute a predictive model.
As an introduction, you can see some of our webinars for more information about process mining.
Analysis done in BICI will replay each process instance using its history and compute the remaining time on each state it went through.
Then percentiles of the remaining time for each of these states will be stored in a prediction model.
The prediction model allows to predict the percentage of chance to be within a target duration.
Then the state of the case is deducted from this percentage and target duration, possible states are on time, late, or predicted late.