Mining logfiles and generating process models is one of the strengths of the ProM Framework. With the differencegraph plugin we introduce a plugin to compare such process models within ProM. 

 

Our plugin uses the instance traffic stored within log files for calculating differences and commonalities of two processes. Therefore we use mining algorithms to generate process models from log files. These process models are used to calculate a new process model called "Differencemodel. Within the difference model each node and edge consist of an marking which describes the changes between the two inputs. Dependent on the input process models the plugin calculates up to five different markings shown in Section See the difference. After calculation the difference model has to be visualized, this visualization is called Differencegraph. 

 

 

A first look at a Differencegraph shows the image below. 

 

The Plugin allows the following inputs: 

Integrated mining algorithms are: 

The Plugin marks differences between the two inputs with three or five colors and symbols. Three colors/symbols (New, Unchanged, Deleted) when the input does not consist of weights (eg. PetriNet). Five colors/symbols when the input consists of weights (eg. HeuristicsNet). 

 

 

Detecting differences between processes can lead to: 

For more information please have a look at the « introduction video».