The latest release is here.
The Performance Spectrum Miner (PSM) is a visual analytics tool for process performance analysis. It takes an event log in a CSV or XES format, visualizes the flow of all cases over time, and gives detailed insights into performance characteristics.
The PSM visualization
- shows how the performance of a process varies over time regarding throughput, volume, steadiness levels, peaks, and drops,
- allows to analyze detailed performance characteristics of each step such as variability in waiting, prioritization of cases, delays and synchronization behavior effecting multiple cases together,
- reveals various performance patterns such as queueing disciplines, batching, prioritization and overtaking, slow movers, temporary bottlenecks, changes in process, and many more, and thereby
- gives insights into different performance variants of the process within each step and across steps, and how these change over time.
The PSM project provides two implementations of the Performance Spectrum Miner as a plugin to the Process Mining Framework ProM and as a stand-alone application.
The PSM project is the result of the joint research project on Process Mining in Logistics between Eindhoven University of Technology and Vanderlande Industries, and developed by Vadim Denisov, Elena Belkina, and Dirk Fahland.
-
- The paper describes a method to extract features for training a performance predicition model from the performance spectrum. The User manual provides details on how to extract performance spectrum-based features. Additional paper-specific documentation and the datasets used in the experiments are documented here.
-
Unbiased, Fine-Grained Description of Processes Performance from Event Data (BPM2018)
-
BPIC'2018: Mining Concept Drift in Performance Spectra of Processes (BPIC2018)
- An OS supporting Java 8, e.g., Microsoft Windows 7 or higher, MacOS, Linux
- 2 GB RAM minimum, 16 GB RAM recommended
- 100MB hard disk space for ProM, 2 GB hard disk space for caches recommended
The PSM is implemented and tested with Java 8 and is not binary compatible with other Java versions.
- Install the most recent JRE/JDK 1.8 64bit
- Make sure that a correct installation of Java is configured: execute
java -version
in the command line. You should get a response like this:
java version "1.8.0_171"
Java(TM) SE Runtime Environment (build 1.8.0_171-b11)
Java HotSpot(TM) 64-Bit Server VM (build 25.171-b11, mixed mode)
*If you do not want to change your current Java installation to Java 8, you can download Java 8 and explicitly call it while starting the PSM or ProM (in 'ProM.bat'), for example://
"C:\Program Files\Java\jre1.8.0_171\bin\java.exe" -jar perf_spec-assembly-1.0.2.jar
- Download ProM nightly build. The PSM has been tested with versions 14th August 2018 and later.
- Run ProM Package Manager (execute
PackageManager.bat
) - Go to tab 'Not installed', find and install plugin PerformanceSpectrum.
- Exit ProM Package Manager
- Recommended for large datasets: open file
ProM.bat
in any text editor and change parameter–Xmx
from4
to a value equal to your laptop's RAM size minus 2 - Execute
ProM.bat
to run the PSM
- Close ProM (if opened) and run ProM Package Manager (execute
PackageManager.bat
) - Go to tab 'Out of date', find and update plugin PerformanceSpectrum.
- Exit ProM Package Manager
- Download and unzip perf_spec-assembly-1.1.1.jar
- Execute
java -jar perf_spec-assembly-1.1.1.jar
to run the PSM
Analyzing the Performance Spectrum of a process with the PSM has three steps:
- Transformation of an event log, either in XES or CSV format, to a Performance Spectrum disk (file) representation. Different performance classifiers and aggregation functions can be used.
- Opening the transformed data for analysis with the PSM
- Exploring the Performance Spectrum
Additionally the PSM allows encoding and exporting performance spectrum-based features into a training and test sets.
- Load the event log into ProM via the Import... button.
- Go to the Action Tab and select the Performance Spectrum Miner plugin from the action list.
- Choose parameters for generating the performance spectrum data.
- A configuration dialog will show providing default values for the transformation.
- See the User Manual for details.
- The transformed data will be stored on disk in the Intermediate storage directory together with a meta-data file (
session.psm
). You can load this transformed data also later into ProM by loading thesession.psm
meta-data file. - Choose Process & open
- The transformation may require some time and main memory depending on the Bin size chosen. Transformation for larger bin sizes are faster and require less memory.
- In ProM the transformed data will then also be opened automatically
- Load the event log (XES format) via the Open... button.
- Choose parameters for generating the performance spectrum data.
- A configuration dialog will show providing default values for the transformation.
- See the User Manual for details.
- The transformed data will be stored on disk in the Intermediate storage directory together with a meta-data file (
session.psm
). You can load this transformed data also later via the Open... button. - Choose Process & open
- The transformation may require some time and main memory depending on the Bin size chosen. Transformation for larger bin sizes are faster and require less memory.
Often event data are available in the CSV format as a database or a distributed file storage dump, stored in one or many CSV files. Converting such dumps to XES format can be difficult for large event logs. The PSM supports a direct import of one or many CSV files. To prepare CSV file(s) for import, put the file(s) into a directory and provide a description as a text ini file with extension .csvdir
. This file must include the following fields (see example):
Field | Sample value | Comment |
---|---|---|
dateFormat |
dd-MM-yyyy HH:mm:ss.SSS |
Datetime format in Java DateTimeFormatter format |
zoneId |
Europe/Amsterdam |
Time zone ID in Java ZoneId format |
startTime |
01-09-2018 00:00:00.000 |
Since then the performance spectrum should be computed, in the format described above |
endTime |
08-09-2018 00:00:00.000 |
Until then the performance spectrum should be computed, in the format described above |
caseIdColumn |
CaseID |
Column name for case ID |
activityColumn |
Activity |
Column name for activity |
timestampColumn |
Timestamp |
Column name for timestamp |
Import your .csvdir
file and use it in the PSM plugin, exactly as XES files.
Import your .csvdir
file and use it exactly as XES files.
Sometimes event data are availabe not as an event log but as segments. Find here how to import such data into the PSM.
- By choosing Process & open during data transformation, the transformed data will be opened automatically. Alternatively, you can also load a previously transformed data set by opening the .psm meta-data file (via Import in ProM, and choosing Performance Spectrum Miner View, or via Open in the stand-alone version).
- Choose parameters for opening. For now use the default values provided, see the User Manual for details.
The main window of the Performance Spectrum Miner is divided into
- a panel visualizing the performance spectrum of the event long
- a control and filtering panel at the bottom that particular contains one sliders to scroll horizontally and two sliders to zoom vertically and horizontally
In the visualization panel, each horizontal segment shows how cases move over time (x-axis) from one activity to the next activity (y-axis).
By default the visualization shows Lines. In the figure below, each colored line describes one case moving from Send Fine to Insert Fine Notification. The x-coordinates of the start and end point of each line visualize the moments in time when Send Fine and Insert Fine Notification occurred, respectively. The color of the line depends on the classification that was chosen in the transformation step, which can be retrieved via the Legend button in the control and filtering panel.
The performance spectrum shows among other things:
- There are cases that are processed very fast (near vertical dark-blue lines) and there are cases processed much slower (sloped lines in light-blue, yellow, and orange).
- The slower cases all have in common that Send Fine occurred for them together with many other cases (at the same moment in time) in a batch, whereas Insert Fine Notification happened individually for each case.
- Batching for Send Fine occurs at irregular intervals and the amount of cases per batch varies greatly over time.
While the Lines show the speed of cases, the amount of cases over time can be visualized by checking Bars in the control and filtering panel.
The stacked bars provide aggreate information about how many cases started, ended, or were pending in particular time-window between the two activities of the segment. The parameters of this aggregation are chosen in the transformation step, see the User Manual for details. In the example above, the stacked bars show that the process experienced a very high amount of cases going from Send Fine to Insert Fine Notification in particular period (the exact time will be shown on the bottom left when hovering the mouse over the respective part of the visualization). The coloring indicates that in this period, the cases were processed much slower than in other period. The number 2988 in the label of the segment tells that there were at a maximum 2988 cases transitioning together through this part of the process.
The Performance Spectrum can be explored in various ways.
- The Options... dialog allows to
- Filter out segments whose throughput of cases (starting/pending/ending cases) in one time window is below/above a certain threshold, e.g, to focus on the most frequently used segments in a process.
- Filtering in and out segments by activity names, using regular expressions, for example
.*Send Fine.*
will include all segments involving Send FineSend Fine:.*
will include all segments starting with Send Fine and leading to some other activity
- see the User Manual for details
- Clicking the left mouse button on the visualization panel shows a context menu with the classes of the classification chosen in the transformation step. Selecting one of them shows only the cases of this class, e.g., only cases whose performance is in the 1st quartile of each segment.
- Right-clicking and dragging a selection box around cases in one segment allows to highlight the selected cases in all other segments (the non-selected cases will be shown in grey). The Clear button in the control panel removes this selection.
More detailed information can be found in
- the User Manual
- Vadim Denisov, Dirk Fahland, Wil M. P. van der Aalst: Unbiased, Fine-Grained Description of Processes Performance from Event Data. BPM 2018: 139-157 (https://doi.org/10.1007/978-3-319-98648-7_9)
- in other related papers (see Sec. Publications above)
The Performance Spectrum Miner project is the result of the joint research project on Process Mining in Logistics between Eindhoven University of Technology and Vanderlande Industries. It was developed by
- Vadim Denisov,
- Elena Belkina,
- Dirk Fahland, Eindhoven University of Technology
The project makes the Performance Spectrum Miner available under the GNU LGPL v3.0 (see file LICENSE)
The objective of the PSM project is to
- demonstrate novel insights into non-steady state, time-variable performance of process event data,
- spur and trigger further research on more detailed process performance analysis, and
- invite collaboration on further developing analysis techniques on the performance spectrum.
Do you have...
- Interesting insights into processes using the Performance Spectrum? Let us know!
- A success story using the Performance Spectrum? Let us know!
- Developed an extension or new feature for the PSM? See the contribution guidelines for this project
More information about the Process Mining in Logistics project focusing on process performance analysis can be found at the project website.
- Find here information on how to build the project from sources.
- Adding custom classifiers is described here.
Please see the contribution guidelines for this project