org.processmining.uncertaintymarker.algorithms

Class AlignedStepDiscovery

java.lang.Object

org.processmining.uncertaintymarker.algorithms.AlignedStepDiscovery

public class AlignedStepDiscovery
extends java.lang.Object

The process discovery step is the third out of four steps performed by this plugin. It is performed after TraceAlignment and before NetAttachment.

This class implements a simple Process Discovery Algorithm which I developed on my own. It works only with an aligned log as input. Beginning with n=1, the algorithm looks at the n-th element of each trace, and groups them together into objects as TransitionData. If the frequency of a certain event is above a certain threshold that is set by the user as a parameter, the event will be included into the resulting Petrinet. Each element in such column n will get an arc from the same predecessor place, and one to the same successor place. This way the resulting net is a sequence of XOR-Connections.

Author:

Oliver

Constructor Summary

Constructors

Constructor and Description
AlignedStepDiscovery()

Method Summary

Modifier and Type	Method and Description
org.processmining.models.graphbased.directed.petrinet.Petrinet	discoverPetrinet(org.deckfour.xes.model.XLog alignedLog, double inclusionThreshold, org.processmining.framework.plugin.PluginContext context) Main method of this class.
void	setColumnDistribution(int column) Aggregate/Summarize all events contained in the specified column into objects of type TransitionData and store them in the transitions list.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

AlignedStepDiscovery

public AlignedStepDiscovery()

Method Detail

discoverPetrinet

public org.processmining.models.graphbased.directed.petrinet.Petrinet discoverPetrinet(org.deckfour.xes.model.XLog alignedLog,
                                                                                       double inclusionThreshold,
                                                                                       org.processmining.framework.plugin.PluginContext context)

Main method of this class. The algorithm produces a Petrinet containing a sequence of XOR-connections of the events in the columns. For each column in the alignedLog, either one, or multiple events are included in the net in an XOR-Connection, depending on whether their relative frequency in this column is above the inclusionThreshold-parameter set by the user. If there is only the gap-sign event in a column which is meeting the condition, this column is skipped.

setColumnDistribution

public void setColumnDistribution(int column)

Aggregate/Summarize all events contained in the specified column into objects of type TransitionData and store them in the transitions list.