Value Series (csv)

The value series (csv) contains up to 9 different numeric value series in a vertical representations. The column headers should be "series1", "series2" ... up to "series9". The numeric values are in the following rows (e.g. 15.3). The cell delimiter is semicolon. The row delimiter is \n. An example would be:

series1series2series3
1.2315.61.5
5.842.751.89
70.53135.55.3

Similarity Matrix (csv)

A Similarity Matrix (csv) contains similarity values between 0 and 100 between process models. The row and column headers contain the model names. Thus, the first column of the first row is empty. The cell delimiter is a semicolon. An example would be:

Model AModel BModel C
Model A10015.370.6
Model B15.310050.89
Model C70.650.89100

Feature Vectors (csv)

The feature vectors csv describes the occurence of activity nodes in particular models. The row header contains all activity labels of all models, the column header contains the models. The values are either 0 (activity is not the model) or 1 (activity is in the model). Line delimiter is \n and cell delimiter is semicolon. An example would be:

Node 1Node 2Node 3
Model A101
Model B111
Model C011

Log Feature Vectors (csv)

The feature vectors csv describes the occurence of an activity (resp. events) in particular traces. The row header contains all activity labels of all models, the column header contains the models. The values are either 0 (activity is not the trace) or 1 (activity is in the trace). Alternatively, the number of occurences can be counted. Additionally, if more than one log is selected as input, the source file for each trace can be optionally presented in the first column. Line delimiter is \n and cell delimiter is semicolon. An example would be:

Event 1Event 2Event 3
(file 1)Trace A101
(file 1)Trace B111
(file 2)Trace C011

Process Log (mxml)

The MXML format is used for describing process logs. Details and the specification of the format can be found here: http://www.processmining.org/logs/mxml

Matching (rdf)

Details on the rdf format used for process matching can found on the site of the Process Matching Contest 2015: https://ai.wu.ac.at/emisa2015/contest.php

Matching (xml)

The datatype xml represents a matching between models. The XML format for saving matchings allows persisting single or multiple matchings of arbitrary arity and complexity. The xml document is structured as follows:

<?xml version="1.0" encoding="UTF-8"?>
  <matchings>
      <matching></matching>
  </matching>
  . . .
</matchings>

Each document has a root element called "matchings. The children of this root element are matching elements. The root element has only elements of type matching as children. A matching element is structured as follows:

<matching name="my_matching">
  <epcs>
  . . .
  </epcs>
  <matches>
  . . .
  </matches>
</matching>

A matching element has a name attribute. The name attribute is expected to contain an arbitrary valid string value (see XML specication http://www.w3.org/TR/REC-xml/). A matching element has two children <epcs> and <matches>. The <epcs> element can contain several <epc> elements representing the models involved in a matching. An <epc> element has an attribute epcDescription that represents a specic model. The <epcs> element is structured as follows:<br><br>
<epcs>
  <epc epcDescription="MODEL_A" />
  <epc epcDescription="MODEL_B" />
  . . .
</epcs>

The <matches> element contains several <match> elements representing a matching's matches. The elements are structured es follows:

<match refEpc=EPK_A value="0.5" status="OPEN" interpretation="SPECIALISATION">
  <relationalProperties> <property name="NON_SYMMETRIC"/>
  . . .
  </relationalProperties>
  <node nodeDescription="Antrag prüfen" epcDescription="EPK_A>
  <node nodeDescription="Antrag technisch prüfen" epcDescription="EPK_B>
  <node nodeDescription="Antrag kaufmännisch prüfen" epcDescription="EPK_B/>
  . . .
</match>

A <match> element has four optional attributes. The attribute 'refEpc' references an epc specied in an <epc> element. 'refEpc' is expected to describe a string value that matches an epcDescription value of an <epc> element. This attribute is used to represent 1:N matches. The attribute 'value' describes a value that can be assigned to a match. 'value' is expected to describe a double value as dened in the section for primitive datatypes. This attribute is used to assign a similarity value to a match. The attribute 'status' describes the state of a match. A match can be either open or closed. Hence the attribute 'status' expects the string value 'OPEN' or 'CLOSED'. The attribute 'interpretation' describes the semantic relation between the nodes. Supported interpretation values can be found in the description of the command 'MATCH'. The <match> element contains several <node> elements that describe the nodes involved in a match. Each <node> element has an attribute 'nodeDescription' that is expected to contain the nodes label/description and an attribute 'epcDescription' which is expected to describe a string value that matches an epcDescription value of an <epc> element.

Example:

<?xml version="1.0" encoding="UTF-8"?>
<matchings>
  <matching name="myMatching">
    <epcs>
      <epc epcDescription="EPK_A" />
      <epc epcDescription="EPK_B" />
    </epcs>
    <matches>
      <match refEpc=EPK_A value="0.5" status="OPEN" interpretation="SPECIALISATION">
        <relationalProperties>
          <property name="NON_SYMMETRIC"/>
        </relationalProperties>
        <node nodeDescription="Antrag prüfen" epcDescription="EPK_A>
        <node nodeDescription="Antrag technisch prüfen" epcDescription="EPK_B>
        <node nodeDescription="Antrag kaufmännisch prüfen" epcDescription="EPK_B/>
      </match>
    </matches>
  </matching>
</matchings>