About
The EDImine project is jointly conducted by the Vienna University of Technology and the Eindhoven University of Technology. EDImine is funded by the Vienna Science and Technology Fund (Wiener Wissenschafts-, Forschungs- und Technologiefonds, WWTF - http://www.wwtf.at). The project is about inter-organizational business process mining in the context of Electronic Data Interchange (EDI).
What is EDI? --> Electronic Data Interchange (Wikipedia).
What is process mining? --> Process mining (Wikipedia).
More information about the project can be found at the project's website at the funding institution.
News
April 1, 2014. We just completed a new case study on inter-organizational business performance analysis together with a beverage manufacturing company. The results will be published soon. Stay tuned!
On April 11, 2014 project member R. Engel is going to have his PhD defense (see the abstract) at 3pm in the meeting room of the Dean's Office of the TU Vienna Faculty of Informatics. This PhD defense is open to the general public.December 1, 2013. The EDImine project is now officially over. However, there are still some results and publications in the pipeline. In particular, we are looking forward to presenting software tools for inter-organizational business performance analyses soon, so stay tuned.
April 30, 2013. We will give a live demonstration of our new EDIminer toolset for process mining from EDI messages at the CAiSE'13 conference (June 17-21, 2013, Valencia, Spain).
Mar 16, 2012. Using the Edimine Correlation Plug-in, we performed a case study on mining inter-organizational business processes from EDI messages together with an automotive supplier company. The results will be published at the CAiSE'12 conference.
Nov 18, 2011. EDImine project members Robert Engel and Marco Zapletal have been successful in entering the finale of the Audi Production Award 2011 held in Ingolstadt from Nov 15-16, 2011 with their logistics-related concept entitled „Supply Chain Analysis“. The concept emerged in the context of EDImine and was selected as one of eight finalists from 68 submissions total with ideas and concepts on manufacturing, production elements or technological processes.
Publications
| 1 | R. Engel, W. Krathu, M. Zapletal, C. Pichler, W.M.P. van der Aalst, H. Werthner: Process Mining for Electronic Data Interchange. In 12th International Conference on Electronic Commerce and Web Technologies (EC-Web 2011), Toulouse, France; August 29 - September 2, 2011, volume 85 of Lecture Notes in Business Information Processing (LNBIP), pp. 77-88. Springer, 2011. [ BibTex ] |
| 2 | R. Engel: A Query Language Enabling Inter-organizational Business Process Mining for EDIFACT Documents. Talk: 9th International Conference on Business Process Management (BPM 2011) - PhD Symposium, Clermont-Ferrand, France; August 28 - September 2, 2011. |
| 3 | C. Pichler, R. Engel, C. Huemer: Validation of Business Document Types based on Feature Models. In 6th International Workshop on Variability Modeling of Software-intensive Systems (VaMoS 2012), Leipzig, Germany, January 25 - 27, 2012; pp. 27-36. ACM, 2012. |
| 4 | D. Jansen: EDI event logs from the EDImine Correlation plugin: An analysis of existing ProM plugins. Co-supervised by W.M.P. van der Aalst and R. Engel, Dec 2011. |
| 5 | W. Krathu, C. Pichler, M. Zapletal, and H. Werthner: Semantic Inter-organizational Performance Analysis using the Balanced Scorecard Methodology. In 35th Jubilee International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO 2012), Opatija/Abbazia, Croatia, May 21-25, 2012, IEEE, 2012. |
| 6 | R. Engel, W.M.P. van der Aalst, M. Zapletal, C. Pichler, and H. Werthner: Mining Inter-organizational Business Process Models from EDI Messages: A Case Study from the Automotive Sector. In 24th International Conference on Advanced Information Systems Engineering (CAiSE 2012), Gdansk, Poland, June 25-29, 2012, volume 7328 of Lecture Notes in Computer Science (LNCS), pp. 222-237. Springer, 2012. |
| 7 | R. Engel, C. Pichler, M. Zapletal, W. Krathu and H. Werthner: From Encoded EDIFACT Messages to Business Concepts Using Semantic Annotations. In 14th IEEE International Conference on Commerce and Enterprise Computing (CEC 2012), Hang-Zhou, China; September 9-11, 2012; IEEE, 2012. |
| 8 | W. Krathu, C. Pichler, R. Engel, M. Zapletal, and H. Werthner: Semantic Interpretation of UN/EDIFACT Messages for Evaluating Inter-organizational Relationships. In 5th International Conference on Advances in Information Technology (IAIT'12), Bangkok, Thailand; December 6-7, 2012; pp. 81-93. Springer, 2012. |
| 9 | R. Engel, J. Rantham Prabhakara, C. Pichler, M. Zapletal, and H. Werthner: EDIminer: A Toolset for Process Mining from EDI Messages. CAiSE'13 Forum at the 25th International Conference on Advanced Information Systems Engineering, Valencia, Spain; June 17-21, 2013; volume 998 of CEUR Workshop Proceedings, pp. 146-153, CEUR-WS.org, 2013. |
| 10 | W. Krathu, R. Engel, C. Pichler, M. Zapletal and H. Werthner: Identifying Inter-organizational Key Performance Indicators From UN/EDIFACT Messages. 15th IEEE Conference on Business Informatics (CBI 2013), Vienna, Austria; July 15-18, 2013, pp. 276–283. IEEE, 2013. |
| 11 | C. Pichler, C. Huemer, M. Wimmer: Evolution of Business Documents based on UN/CEFACT's Core Components. Accepted for International Journal of Software and Informatics (IJSI), 7(2013), 2; pp. 331-356. 2013. |
| 12 | R. Engel, J. Rantham Prabhakara, C. Pichler, M. Zapletal, and H. Werthner: Inter-organizational Business Processes: On Redundancies in Document Exchanges. Accepted for 15th International Conference on Electronic Commerce (ICEC 2013), Turku, Finland; August 13-15, in: Effective, Agile and Trusted eServices Co-creation, number 19 in TUCS Lecture Notes, pages 51–66. Turku Centre for Computer Science, 2013. |
| 13 | R. Engel, W. Krathu, C. Pichler, M. Zapletal, and H. Werthner: Towards EDI-Based Business Activity Monitoring. Workshop on Methodologies for Robustness Injection into Business Processes (MRI-BP'13) at the 17th IEEE International EDOC Conference (EDOC 2013), Vancouver, Canada; September 9-13, 2013; In In: 17th IEEE International Enterprise Distributed Object Computing Conference Workshops (EDOCW 2013), pages 158–162, 2013. |
| 14 | R. Engel, J. Rantham Prabhakara: A Case Study on Analyzing Inter-organizational Business Processes from EDI Messages using Physical Activity Mining. In 47th Hawaii International Conference on System Sciences (HICSS 2014), Waikoloa Village, Hawaii, USA, January 6-9, 2014. IEEE, 2014. |
| 15 | W. Krathu, C. Pichler, R. Engel, M. Zapletal, H. Werthner, and C. Huemer: A Framework for Inter-organizational Performance Analysis from EDI Messages. In 16th IEEE Conference on Business Informatics (CBI 2014), Geneva, Switzerland, July 14-17, 2014. IEEE, 2014. |
| 16 | W. Krathu, G. Xiao, C. Pichler, H. Werthner, J. Neidhardt, M. Zapletal, and C. Huemer: Inter-organizational Success Factors: A Cause and Effect Model. Accepted and to be appeared in Information Systems and e-Business Management, Springer. |
EDI Data Preprocessing Tools
Several tools are developed in the course of the EDImine project for EDI data preprocessing. This includes:
- The EDIminer toolset, a stand-alone application, for semantic preprocessing EDI data by generating EDI ontologies and knowledge base from EDI messages as well as generating events by mapping EDI data to events.
- The BI preprocessing tool, a stand-alone application, also for semantic preprocessing EDI data by generating business information (BI) ontologies on top of the EDI ontologies.
- The Edimine Correlation plug-in, a ProM plug-in, for correlating events into corresponding cases or processes.
EDIminer
EDIminer is our new toolset for process mining from EDI messages and a main result of the EDImine project. It is currently implemented as a Java stand-alone application. With regard to its functionality, it fully supersedes the Edimine Correlation Plug-in for ProM 6. Please find a detailed description of the tool in this conference paper.
The tool is not quite ready yet to be fully released to the public; however, if you want to try an early release, you can download it here: EDIminer.jar.
We intend to release a mature version of the tool into the public domain soon. Stay tuned.
BI Processing Tool
To conceptualize EDI data into generic business information concepts, the business information (BI) concepts are generated and mapped to EDI data in the EDI ontologies. The BI preprocessing tool is developed as a stand-alone java application for supporting this task. Currently, it provides the default mappings of ORDERS and INVOIC message type of the UN/EDIFACT releases: d96a, d97a, and d01b. However, it also allows for any user-defined mappings which follow our template (see this conference paper for details, and the mapping (owl) file as for examples). You can download the BI preprocessing tool here: EDImineBiprocessor.jar. Please note that it requires DLV system for reasoning engine which can be downloaded here.
Edimine Correlation Plug-in for ProM 6
One of the outcomes of the EDImine project is the Edimine Correlation Plug-in for ProM 6. It is publicly available. It was originally intended as a proof-of-concept implementation in a research project, so do not expect any bells and whistles.
The plug-in's purpose is to generate XES event logs from sets of EDI messages. This enables the subsequent application of common process mining algorithms on the data.
License
The plug-in is licensed under the terms of AGPL Version 3.
The distribution bundles the following libraries:
| Library | License | Remarks | |
|---|---|---|---|
| Smooks 1.4 | LGPL | used for EDI-->XML transformation | |
| Jena 2.6.4 | BSD-style | used for reading the Edimine Ontology | |
| RSyntaxTextArea 1.5.1 | LGPL | used for nicely visualizing EDI/XML data |
Installation
To install the plug-in, get a nightly build of ProM 6 and install the EDIMine package from the Package Manager.
Usage
- Currently the plug-in works only with EDIFACT data.
- The plug-in can read EDIFACT data from a folder. Each file in this folder is assumed to contain an EDIFACT interchange. Files that cannot be read for some reason are skipped.
- At the moment, the following UN/EDIFACT releases are supported out-of-the-box: d96a, d96b, d97a, d97b, d98a, d98b, d99a, d99b. If you need another release, install the corresponding Smooks mappings to your classpath (pick the version 1.4 releases of the mappings). Then create a custom Smooks configuration file for the plug-in (derive it from the plug-in's default Smooks configuration as shown in the GUI). In case of difficulties you can try asking this person for support.
- The plug-in is capable of adding business information from the EDIFACT data to the generated XES log. The mappings between the raw EDI data and conceptualized entries in the XES event log are stored in the Edimine ontology. You can use a custom ontology with the plug-in (derived from the plug-in's internal ontology) in order to include additional business information. Again, in case of difficulties you can try asking this person for support.
- Hint: Double-clicking on a message in the visualization pane of the plug-in will open the XML-ized version of the message.
Tuning
You can extremely speed up (and it is highly recommended to do so) the Edimine Correlation Plug-in by having the Saxon library on your classpath. Once it is on your classpath, it will be automatically used by the plug-in for XML/XPath processing. Saxon is not included in the plug-in's package for licensing reasons. Just download it and put in on ProM's classpath.Business Performance Analysis Tool
In addressing the evaluation of inter-organizational performance from EDI messages, the EDImine BSC Framework is developed. The core contribution of the framework is to integrate top-down and bottom-up performance analysis approaches. In particular, it enables (i) bottom-up KPI definitions and calculations based on the results of process minings (i.e., process models and event log) as well as business information contained in EDI messages, and (ii) top-down alignment of KPIs and business strategies by leveraging the Balanced Scorecard (BSC) method. Detail on the framework is currently under publication process and we will announce as soon as it is published.
The framework is relaized as a plug-in of ProM 6, namely EDImine (Flex Based) BSC plug-in.
EDImine BSC Plug-in for ProM 6
The EDImine (Flex Based) BSC plug-in is a prototype developed as a proof-of-concept implemenation of the framework. It takes event log, mined process model, and EDI & BI (business information) ontologies as inputs for defining/modeling and calculating KPIs. The plug-in is contained in the EDIMine package of ProM 6.The plug-in includes (i) BSC ontology and (ii) the knowledge base of some predefined success factors and KPIs. Note that currently the knowledge base is not yet complete, only some KPIs are currently predefined. Full set of our identified KPIs (see this conference paper for the identified KPIs) will be added in the knowledge base soon. However, the knowledge base is used only for KPI suggestion. Therefore, the plug-in should work since there is no effect to the main functions of the plug-in (i.e., the BSC implementaion including modeling business objectives, success factors, and KPIs as well as BSC calculation).
The plug-in works with other additional plug-ins and tools as follows:
- DLV system for reasoning engine which can be downloaded here.
- EDImine Event Class and BI Mapper plug-in for creating mapping between event types to business information in ontologies. It is already included in our EDIMine ProM package.
- EDImine Flex-based Log Filter plug-in (optional) for filtering event logs by event-related, or business information. It is already included in our EDIMine ProM package.
License
The plug-in is licensed under the terms of AGPL Version 3.
The distribution bundles the following libraries:
| Library | License | Remarks | |
|---|---|---|---|
| DreW 0.3-beta | Apache | used for reasoning and querying over ontologies | |
| Pellet 2.3.0 | AGPL Version 3 | used for reasoning over ontologies | |
| OWL API 3.4 | Apache | used for manipulating OWL ontologies | |
| exp4j 0.3.10 | Apache | used for modeling and calculating the algebraic expressions of KPIs |
Installation
The plug-in is added to the EDIMine package. However, the current public version of the EDIMine package does not yet include this plug-in. If you want to use it, you can obtain source code from ProM SVN (https://svn.win.tue.nl/repos/prom). One option to run the latest version of ProM 6 plug-ins is to use them via Eclipse. You need Eclipse and Subclipse, then follow these steps:
- Create the SVN project by checking out projects from ProM SVN
- Create project repository location to ProM SVN (i.e., https://svn.win.tue.nl/repos/prom). Inside the repository, there is a folder named "Packages" which contains several packages including our EDIMine package.
- Browse in the EDIMine package and select folder "Trunk" for checking out the package as your SVN project.
- To run the package, select file named "ProM with UITopia (EDIMine).launch" and run as "ProM with UITopia".
Usage
- The plug-in currently supports flexible models. Currently, there are existing plug-ins available in ProM that convert other type of models (e.g., hueristic model, or petri net) to flexible models.
- The plug-in takes three main inputs: event log, related process model, and EDI & BI ontologies (obtained from the EDIminer toolset and the BI proprocessing tool).
- Event log can be filtered by either our provided EDImine Flex-based Log Filter plug-in or other existing ProM plug-in before used as an input.
- The the mappings between event types and business information concepts are required before using the plug-in. This can be obtained from the EDImine Event Class and BI Mapper plug-in which is already included in the EDIMine package.
- The plug-in provides a cockpit for implementing business performance anaysis through the BSC method. It has a KPI panel provided for defining and modeling KPIs from input data (i.e., event log, process model and EDI & BI ontologies). It also provides a business objective panel for modeling business objectives, success factors as well as connected them to corresponding KPIs according to the BSC method. Based on the scorecard modelded in the cockpit, the plug-in calculate the performance scores and visualized as tables.
- If you have some difficulties on the installation and the implemenation of the plug-in, please contact this person.
Support the EDImine Project
If you were successful in analyzing your EDI data with our plug-in, please consider supporting the EDImine project by providing us with sample data.
Institute of Software Technology and Interactive Systems
Vienna University of Technology
Favoritenstrasse 9-11/188
1040 Vienna
Austria
