Process Repositories
Important Dates
11 April 2012 - opening presentation
(Slides)
15 April 2012 - topic submission
16 April 2012 - topic assignment notification
25 April 2012 - lecture: how to make a scientific presentation
2 May 2012 - short presentations
9 May 2012 - lecture: how to write a scientific paper
6 June 2012 - technical presentation
| Schedule for technical presentations |
|---|
| Wednesday June 6, 2012 | room |
|---|
| 11:00-11:30 Uhr |
1. presentation |
A-1.2 |
| 11:30-12:00 Uhr |
2. presentation |
| 12:00-12:50 Uhr |
lunch | mensa |
| 12:50-13:20 Uhr |
3. presentation |
H-E.52 |
| 13:20-13:50 Uhr |
4. presentation |
| 13:50-14:00 Uhr |
short break |
| 14:00-14:30 Uhr |
5. presentation |
| 14:30-15:00 Uhr |
6. presentation |
| Friday June 8, 2012 | |
| 10:00-10:30 Uhr |
7. presentation |
H-2.58 |
| 10:30-11:00 Uhr |
8. presentation |
| 11:00-11:30 Uhr |
9. presentation |
19 June 2012 - paper draft submission
20 June 2012 - paper draft distribution
27 June 2012 - review submission
28 June 2012 - review distribution
4 July 2012 - final presentation
22 July 2012 -
final paper & prototype submission
All the submission deadlines are 23:59 CET.
Description
In an endeavor to improve the performance of a company, business processes play an important role. (Graphical) models of business processes are an effective tool to communicate, document, and continuously improve processes. Organizing and maintaining these valuable assets of a company can get a very complex task and understanding the relations between the models becomes more and more difficult with the size of the model collection.
If the business process model collection reaches a certain size, so called business process repositories are required to effectively store and maintain it. Closely related is the question, as to how to relate and present the models to the business analyst. Different process architectures were proposed and serve different purposes.
In this seminar the master students have the opportunity to make their contribution to this emerging field of business process management.
The
topics include, but are not limited to:
Topic Proposals
- Evellin Cardoso
- Topic (1): Model-driven approach implementation of KPIs for Process Models (Co-supervisor: Andreas Rogge-Solti)
- Context: BPMN is the current standard for business process modeling, providing both a notation for modeling processes as well as a semantics for execution. A trend nowadays is the monitoring of process performance by using Key Performance Indicators (KPIs).
- Task:
- Find a use-case in which the performance of business processes must be measured in terms of KPIs, such as duration of activities and/or processes,
- Implement this use case in a model-driven way using the language for KPI modeling proposed in (del-Río-ortega et al, 2011)(first reference) as input and Java code for the activity engine as output,
- This model-driven implementation should generate Java code. Make tests on the basis of the selected case-study and afterwards analyze the obtained results.
- Literature:
- del-Río-ortega, Adela, Resinas, Manuel e Ruiz-Cortés, Antonio. 2010. Defining Process Performance Indicators: An Ontological Approach. Hersonissos, Crete, Greece : Springer-Verlag, 2010. pp. 555-572.
- BPMN Method and Style
- Rami-Habib Eid-Sabbagh
- Topic (2): BPM Process Model Translater
- Context: Many organizations have dependencies in multiple countries and work across the world. When modeling processes the business process experts often use the local language, so that the process models cannot be understood by the colleagues in other countries who do not speak this particular language.
The language used in process models is usually very specific. Hence, a good translation, taking into account the context, is of major importance. To overcome this problem and fasten the distribution of process knowledge among countries, a translation tool can make a significant contribution. Hence, a tool that translates activities and event labels of processes German to English and vice versa is needed.
The aim of the task is to provide a correct and accurate translator for labels that takes context of the process models in to account:
- Task:
- Research and evaluate: translation possibilities based on machine translation technologies (e.g. MOSES)
- You may also consider other translators as Google or BING
- Detect various naming styles and decompose labels for step by step translation (provided by Henrik)
- Develop algorithms to use the context of activities and events for the translation ("shot" has a different meaning with the context "vaccine" as with the context "gun").
- Translate the decomposed parts of the labels using the context
- Implement and evaluate your approach
- Scope: German – English and roundtrip
- Literature:
- http://www.statmt.org/moses/
- Leopold, H., Smirnov, S., Mendling, J.: On the refactoring of activity labels in business process models. Information Systems (IS) 37(5) (2012) 443 – 459
- Topic (3): Labeling Clusters of Process Collections
- Context: With the prevalence of business process management in the private and public sector, large process collections are created and shift into focus. To be able to harvest the underlying information, process collections need to be made easily accessible providing intuitive navigation and search. The knowledge stored in process collections is often not treasured to the best possible extent. Process collections are semi or structured. Especially for large collections, the need for an inherent and intuitive structure and navigation is of importance for the retrieval of process models. Navigation structures can be provided by using hierarchical agglomerative clustering algorithms that result into a hierarchical tree, a dendrogram. However, creating clusters is just one part of the medal. An important role for navigation plays the labeling of the clusters in the hierarchy tree. There exist several algorithms for labeling clusters, e.g. choosing centroid, or using mutual information. Some of them regard only the local members of the clusters, other more advanced ones also consider all clusters to choose a well-distinguishable label. In hierarchical clustering the labeling comes with another difficulty the hierarchical relation of clusters which needs to be taken into account. To be able to generate good hierarchical navigation structures with useful labels in the domain of process collections a labeling algorithm considering the hierarchical relations needs to be developed.
- Task:
- Research and evaluate cluster labeling algorithms
- Develop algorithm for labeling clusters of process models
- Consider hierarchical relations of clusters and labels
- Implement a prototype
- Evaluate and compare your approach to existing algorithms
- Literature:
- http://www.statmt.org/moses/
- Baeza-Yates, R., Ribeiro-Neto, B.: Modern Information Retrieval. Addison Wesley (1999)
- Manning, C.D., Raghavan, P., Schütze, H.: Introduction to Information Retrieval. Cambridge University Press (2008)
- Muhr, M., Kern, R., Granitzer, M.: Analysis of structural relationships for hierarchical cluster labeling. Proceedings of the 33rd international ACM SIGIR conference on Research and development in information retrieval. pp. 178-185. ACM, New York, NY, USA (2010)
- Treeratpituk, P., Callan, J.: Automatically labeling hierarchical clusters. Proceedings of the 2006 international conference on Digital government research. pp.167-176. ACM, New York, NY, USA (2006)
- van Rijsbergen, C.J.: Information Retrieval. Butterworth (1979)
- Nico Herzberg
- Topic (4) - How to configure event monitoring points
- Context: Today’s organizations strive to evaluate their executed processes with respect to performance and conformance with process monitoring and process analysis tools. Implementing such a capability in process-aware information systems is not an issue, however most processes are still carried out manually (e.g. in the Healthcare domain). In those environments the effects of the process execution are limited to a few data that typically are not related to a process instance. For such environments the BPT Group works on architecture (Architecture.pdf: Process Monitoring Architecture) to bridge the gap between the IT landscape, where the footprints of the process execution could be found widely scattered, and the process monitoring and analysis system.
In the architecture a service-based binding is foreseen which connects the so-called event monitoring points with the IT systems. The binding layer is differentiated into a capturing service layer, an event bus and a correlation layer.
When it comes to the configuration of the event monitoring points which is the correlation to certain events it must be easy to find the right events to correlate.
- Task: Describe a proposal of how a concrete solution could look like for an event repository that could be used for the configuration of event monitoring points. The event repository should hold all information about events that could be tracked in the IT system landscape including the event capturing service that provides the implementation of the data provision. Find out if an SOA could be applied for that requirement. Investigate on topics like creation of a repository entry, search, meta-data and event capturing service binding. Evaluate if the architecture is ready for your approach or need to be adjusted or refined.
The work to be done is based on a rough process model of the evaluation of a patient for liver transplantation. This process will hold three event monitoring points. The IT system will provide data out of three data sources (a database, a spreadsheet and a web service).
- Literature:
- N. Herzberg, M. Kunze, A. Rogge-Solti. Towards process evaluation in non-automated process execution environments. 4th Central-European Workshop on Services and their Composition (ZEUS), 2012
- H.R. Motahari-Nezhad, R. Saint-Paul, F. Casati, and B. Benatallah. Event correlation for process discovery from web service interaction logs. VLDB Journal, 20(3):417–444, 2011.
- Oasis. Reference Model for Service Oriented Architecture. Public Review Draft 2. Oasis Standard, 2006
- David Luckham. The Power of Events – An introduction to complex event processing in distributed enterprise systems; 2010
- Topic (5) - Calculate KPIs out of event monitoring points
- Context: Today’s organizations strive to evaluate their executed processes with respect to performance and conformance. Therefore organizations define key performance indicators (KPI) in alignment with their business goals to monitor and control their business. Measuring KPIs in process execution environments means to measure the execution of various processes itself.
The BPT Group works on architecture that foresees so-called event monitoring points. Those represent points of interest for monitoring in a process that calculation algorithms of KPI can use.
- Task: Find and evaluate existing approaches about generation of calculation algorithms for KPIs. Investigate on what literature covers about this topic already and also have a look at similar topics like service level agreement measurement. Propose one of the existing approaches for the event monitoring point environment or may come up with an own approach. Describe how this approach could fit into the architecture proposed by the BPT group and what may needs to be adjusted or refined.
The work to be done is based on a rough process model of the evaluation of a patient for liver transplantation which holds three event monitoring points. The KPIs could be created according the given scenario, for example: The time between first contact with the transplantation until listing at Eurotransplant should not exceed 1 month.
- Literature:
- N. Herzberg, M. Kunze, A. Rogge-Solti. Towards process evaluation in non-automated process execution environments. 4th Central-European Workshop on Services and their Composition (ZEUS), 2012
- Adela del-Río-Ortega, Manuel Resinas, Antonio Ruiz Cortés: Defining Process Performance Indicators: An Ontological Approach. OTM Conferences (1) 2010: 555-572
- Neely, A.D., Gregory, M. and Platts, K. (1995), “Performance measurement system design: a literature review and research agenda”, International Journal of Operations & Production Management, Vol. 15 No. 4, pp. 80-116.
- Matthias Kunze
- Process Model Alignment (6)
- Context: Any process similarity measure relies on a process alignment, that is, information, which nodes represent the same activity in two processes. Such an alignment is usually based on similarity of phrases, i.e., labels of business process models, but takes also into account the relation of activities to their ancestors. Alignments can be 1:1, n:1, and iteratively be optimized.
- Task: Research existing string similarity and activity alignment algorithms and integrate them into a general framework.
Experimentally compare and evaluate these algorithms.
Incorporate them as a library into the PromniCAT? framework.
- Literature:
- Felix Naumann, Melanie Herschel: An Introduction to Duplicate Detection. Morgan & Claypool Publishers 2010
- Jonathan Gumpp: Die Berechnung der Ähnlichkeit von Geschäftsprozessen. Eine Implementierung. Seminararbeit 2010.
- Matthias Kunze, Matthias Weidlich, Mathias Weske: Efficient Retrieval of Similar Business Processes (Draft).
- existing algorithms and documentation
- Comparison of Model Collections: Complexity and Vocabulary (7)
- Context: For a set of process models created by students, we empirically assessed complexity and used vocabulary by means of diverse process model metrics. these measures shall be applied to a broader set of process model collections and conclusions shall be drawn, on the different aspects these collections expose.
- Task: Import further process model collections into the PromniCAT? framework (requires implementation of importer and parser) and derive complexity and vocabulary measures. Discuss differences and commonalities.
- Literature:
- M. Kunze, A. Luebbe, M. Weidlich, and M. Weske. Towards Understanding Process Modeling – The Case of the BPM Academic Initiative. In R. Dijkman, J. Hofstetter, and J. Koehler, editors, Third International Workshop on Business Process Model and Notation (BPMN 2011), volume 95 of LNBIP, pages 44–58. Springer, 2011.
- J. Mendling. Metrics for Process Models: Empirical Foundations of Verification, Error Prediction, and Guidelines for Correctness, volume 6 of Lecture Notes in Business Information Processing. Springer, 2008.
- M. Z. Muehlen and J. Recker. How much language is enough? theoretical and practical use of the business process modeling notation. In CAiSE? ’08: Proceedings of the 20th international conference on Advanced Information Systems Engineering, pages 465–479, Berlin, Heidelberg, 2008. Springer-Verlag.
- BPMN Conformance Levels and Model Characteristics (8)
- Context: In earlier work, we examined the BPMAI process model collection with regards to complexity and used vocabulary. However, practitioners are interested in other aspects, e.g, which BPMN conformance levels are applied, are models structurable, inherently complex, do they expose certain correctness properties, e.g., soundness, relaxed soundness, weak soundness, ..., and to which model class do they belong.
- Task: Study different literature and existing code (jBPT, PromniCAT?) and develop algorithms in the fashion of analysis modules for PromniCAT? to detect above properties.
- Literature:
- M. Kunze, A. Luebbe, M. Weidlich, and M. Weske. Towards Understanding Process Modeling – The Case of the BPM Academic Initiative. In R. Dijkman, J. Hofstetter, and J. Koehler, editors, Third International Workshop on Business Process Model and Notation (BPMN 2011), volume 95 of LNBIP, pages 44–58. Springer, 2011.
- Object Management Group. Business Process Model and Notation (BPMN) Specification, Version 2.0, 2009.
- W. M. P. van der Aalst. Verification of Workflow Nets. In ICATPN ’97: Proceedings of the 18th International Conference on Application and Theory of Petri Nets, pages 407–426, London, UK, 1997. Springer-Verlag.
- W. van der Aalst, K. van Hee, A. ter Hofstede, N. Sidorova, H. Verbeek, M. Voorhoeve, and M. Wynn. Soundness of Workflow Nets: Classification, Decidability, and Analysis. Technical report, Eindhoven University of Technology, 2008.
- Artem Polyvyanyy, Luciano García-Bañuelos, and Marlon Dumas. Structuring Acyclic Process Models. Proceedings of the 8th International Conference on Business Process Management (BPM). Hoboken, NJ, US, September 2010
- Model Evolution Patterns (9)
- Context: Based on the versions of model in the BPMAI, it is possible to trace the process of model creation and get more insights into the act of process modeling. This can also be linked to the number of people that have contributed to the models. On average, models in the BPMAI collection have 4.27 revisions and 20% have 6 or more revisions. One particular BPMN process model even exposed 80 revisions.
- Task: Evaluate methods to analyze process model revisions towards different aspects and deduce typical patterns of model evolution, e.g., are there typical phases of modeling identifiable, such as initial creation, exception handling, role allocation, etc.
- Literature:
- M. Kunze, A. Luebbe, M. Weidlich, and M. Weske. Towards Understanding Process Modeling – The Case of the BPM Academic Initiative. In R. Dijkman, J. Hofstetter, and J. Koehler, editors, Third International Workshop on Business Process Model and Notation (BPMN 2011), volume 95 of LNBIP, pages 44–58. Springer, 2011.
- tbd
- Andreas Meyer
- Deriving data from control flow information (10)
- Context: Data plays major role in business process management, e.g., for process enactment. But usually data is not modeled explicitly in process models. However, it is often hidden in control flow information – more specifically in activity labels.
- Task: Derive data objects from control flow information. Provide the implementation as extension to the PromniCAT? framework. (Extend with means to derive data models (data structure) from occurrence and usage of data objects on different levels of abstraction within one process repository or domain?)
- Literature:
- National language processing: stanford parser, paper from Henrik Leopold
- Label matching: paper from Marlon Dumas, Sergej Smirnov
- Correction and simplification of process models based on object life cycle information (11)
- Context: A process scenario contains a process model and a set of object life cycles (one for each data object used in the process model). There exist notions to check for correctness between the process model and the object life cycles. One of them is the notion of weak conformance, which checks whether all data state transitions induced by activities of the process model, are comprised by the object life cycle in terms of data state reachability. Process scenarios not satisfying the notion of weak conformance can be corrected in several ways, e.g., by modifying the control flow, deleting affected model artifacts and change the states in the process model as well as adapting the object life cycle by adding new data state transitions. Usually, such modifications are ambiguous and the best fitting correction needs to be chosen. Therefore, human interaction is required. However, the possible modifications can be ranked and ordered to this regard to ease decision taking by the human. Additionally, modification may result in structural changes to the process models, e.g., sequentialization of blocks. Incorporate these model simplifications into your proposed solution.
- Task: Introduce means to handle corrections of process scenarios (and incorporate process model simplification) to achieve satisfaction of the notion of weak conformance.
- Literature:
- A. Meyer, A. Polyvyanyy, M. Weske: Weak Conformance of Process Models with respect to Data Objects
- BPM submission
- J. Küster, L. Ryndina, H. Gall: Generation of Business Process Models for Object Life Cycle Compliance
- K. Ryndina, J. Küster, H. Gall: Consistency of Business Process Models and Object Life Cycles
- A. Awad, G. Decker, N. Lohmann: Diagnosing and Repairing Data Anomalies in Process Models
- S. Sun, J. Zhao, J. Nunamaker, O. Sheng: Formulating the Data-Flow Perspective for Business Process Management
- Correction of BPM-AI process models (12)
- Context: The BPM-AI process model repository contains a large amount of process models. Unfortunately, a decent amount of them contain structural errors, so that they cannot be used for analyses. For instance, tasks and edges are not connected but by utilizing coordinates of both shapes, the connection can be drawn. Alternatively, analysis of the svg/png representation of process model may support the correction procedure. (engineering problem)
- Task: Identify and correct structural errors in process models and implement your solution as modul to the PromniCAT? framework.
- Data and resource metrics for process models (13)
- Context: Several process metrics are around. But these mainly target control flow aspects of process models. However, the data and resource perspectives of process models are also important as these describe who actually performs the tasks (resource perspectives) and what artifacts are required and created in the process and how these artifacts are manipulated over time (data perspective).
- Task: Create and evaluate data and resource metrics for process models. Therefore, implement them as extension to the PromniCAT? framework. (kind of explorative and validation of results may be difficult)
- Literature:
- Andreas Rogge-Solti
- Multi-process-monitoring views (14)
- Context: Monitoring what happens in all the processes of a company is not trivial. This topic is centered around different views on process models. One candidate is the role-based view, which should present the current state of the processes of a role. A person belonging to that role can see, how far the process instances have progressed.
- Task: Browse relevant literature that deals with views of processes, categorize different approaches for multiple process views, e.g., single-person / multiple instances, single role multiple process. Sketch your ideas to provide a multi-process monitoring view given a process model repository with processes containing role information. Evaluate your ideas with a user group.
- Literature:
- D. Schumm, F. Leymann, A. Streule. Process Viewing Patterns. Proceedings of the 14th International IEEE Enterprise Distributed Object Computing Conference (EDOC 2010), IEEE Computer Society Press, 2010.
- D. Grigori, F. Casati, M. Castellanos, U. Dayal, M. Sayal, and M.C. Shan. Business Process Intelligence. Computers in Industry, 53(3):321–343, 2004.
