Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Reducing Complexity Of Large EPCs


Published on

Business processes are an important instrument for understanding and improving how companies provide goods and services to customers. Therefore, many companies have documented their business processes well, often in the Event-driven Process Chains (EPC). Unfortunately, in many cases the resulting EPCs are rather complex, so that the overall process logic is hidden in low level process details. This paper proposes abstraction mechanisms for process models that aim to reduce their complexity, while keeping the overall process structure. We assume that functions are marked with efforts and splits are marked with probabilities. This information is used to separate important process parts from less important ones. Real world process models are used to validate the approach.

Published in: Technology
  • Be the first to comment

Reducing Complexity Of Large EPCs

  1. 1. Reducing Complexity of Large EPCs Artem Polyvyanyy Sergey Smirnov Mathias Weske EPK 2008 27 November 2008
  2. 2. Motivation <ul><li>Research project with AOK Brandenburg </li></ul><ul><li>Goal: detailed process models abstract process models </li></ul><ul><li>≈ 4 000 EPCs </li></ul><ul><li>graph-structured process models </li></ul><ul><li>model elements have annotations: </li></ul><ul><ul><li>connections -> probabilities </li></ul></ul><ul><ul><li>functions -> average execution time </li></ul></ul><ul><li>average execution time of a process must be preserved </li></ul>
  3. 3. Motivation <ul><li>Example model : </li></ul><ul><ul><li>> 300 nodes </li></ul></ul><ul><ul><li>>150 functions </li></ul></ul><ul><ul><li>graph-structured model </li></ul></ul>
  4. 4. Extended EPC <ul><li>(E, F, C, A, t, e r , p r ) is an extended EPC </li></ul><ul><ul><li>E is a set of events, E ≠ Ø </li></ul></ul><ul><ul><li>F is a set of functions, F ≠ Ø </li></ul></ul><ul><ul><li>C is a set of connectors </li></ul></ul><ul><ul><li> is a set of nodes </li></ul></ul><ul><ul><li>is a set of connections </li></ul></ul>
  5. 5. Auxiliary Concepts <ul><li>Mean occurrence number of a node </li></ul><ul><li>the mean number of node occurrence in a process instance </li></ul><ul><li>Absolute effort of a function e a </li></ul><ul><li>the mean effort contributed to the execution of the function in </li></ul><ul><li>a process instance </li></ul><ul><li>Process absolute effort e p a </li></ul><ul><li>the mean effort required to execute a process instance </li></ul>
  6. 6. Example 0.92 0.08 10 minutes 9.2 minutes 20 minutes 1.6 minutes transition probability relative effort absolute effort
  7. 7. Business Process Model Abstraction What model elements are insignificant? <ul><li>absolute effort of a function e a </li></ul><ul><li>elementary abstractions </li></ul><ul><li>transformation of a fragment </li></ul><ul><li>abstraction strategy </li></ul><ul><li>organization of elementary </li></ul><ul><li>abstractions </li></ul>How to abstract insignificant elements?
  8. 8. Elementary Abstractions Block Dead end Sequence Loop
  9. 9. Sequence Elementary Abstraction
  10. 10. Block Elementary Abstraction !
  11. 11. Dead End Elementary Abstraction
  12. 12. Loop Elementary Abstraction
  13. 13. Abstraction Example 0.00 0.37 1.00
  14. 14. Conclusions <ul><li>Conceptualized the problem of </li></ul><ul><li>business process model abstraction </li></ul><ul><li>Proposed the abstraction approach based on rules describing </li></ul><ul><li>structural transformations </li></ul><ul><li>dependencies between elements non-functional properties </li></ul><ul><li>Developed the implementation of the approach </li></ul>