The DCR Portal contains an app, DCR Process Discovery, which can analyse flows and suggest rules to the user.
The App is based on latest research from Copenhagen University on process mining and process discovery. Identifying rules in a DCR process models has been a topic for several student project at IT University and Copenhagen University and we're working with researchers to refine the algorithms to make an industrial strength solution.
The App can analyze either Flows or specific log files in XES or DCR XML Log format. The image below illustrates analysis of Flows and recommended rules.
Clicking on the leftmost grey icon will turn it green to add and red to ignore the rule. The App captures users accept or reject of rules which is used to train the algorithm to provide better suggestions. We also safe information about process model, flows and log in order to train the computer.
You can add all the rules in one click without filling out the explanation by clicking CTRL+Add rules button on the same time. Normally we requirement explanations for each rule but you can avoid this simply clicking CTRL too.
Version 1.3 released September 2020 makes it possible to choose which rules to mine:
Choose to mine
- Conditions
- Responses
- Includes
- Excludes
It is also possible to add a milestone for each condition added. We refer to conditions as the "condition/milestone" rule, or a pre-condition.Version 2.0, released December 30th 2020, supports mining Forbidden flows using the Rejection Miner developed by Tijs Slaats and Søren Debois as part of the EcoKnow project. We now show which process mining algorithms that found the business rule, i.e. Rejection for rejection miner, DisCoveR for DisCoveR miner and Both of rule was found by both miners.
