Homework 2: Performance Mining
We consider an event log of a handling loan applications, which can be found here: https://owncloud.ut.ee/owncloud/s/w6adDY7i5PkPo3w.
Upload the given event log to the Apromore server and solve the following tasks (also providing supporting screenshots and explanations):
- [1 point] We have discussed in the lectures that process defects and reworks can ofen be identified by activity names, self-loops, and short-loops. Identify all activities in this process that indicate some kind of defect or rework.
- [1 point] Use different process map overlays to determine which activity from the previous question is likely to have the greatest effect on the cycle time of the process. Provide the reasoning why you picked that activity and compare the cycle time of cases with and without that activity.
- [1 point] This event log contains loans from multiple countries. Which of the activities identified in the first question are more common in Australia (AUS) and which are more common in New Zealand (NZL)?
- [1 point] Every activity in this process belongs into one of four stages. For each stage, identify the average cycle time from the beginning of the stage (first execution of any event of that stage) to the end of the stage (last execution of any event of that stage).
- [1 point] Determine the average processing time and the average waiting time within each stage of the process, excluding the waiting times between the stages and considering only the activities belonging to that stage (i.e., if an activity from a different stage is executed in the middle of the current stage, then, contrary to the previous question, the processing time and waiting time of that activity should be excluded).
- Hint: When importing the event log, use the column mapping 'Event Attribute (Perspective)' to enable a new perspective in the process discovery view.