Applying and extending techniques of data mining to re-discover business process models has becoming popular in recent years. This can help not only finding process models in cases where there were typically no rigorous design for business process (e.g., case management applications), but also finding interesting unknown regularities of process models. The result of process mining can provide very useful information for improving process models and process management. Business process execution typically follow some pre-designed process models. However, process models are usually following by a majority but not all process executions since business processes often encounter exceptional situations where detour to the existing process models are necessary. Periodically, the enterprise needs to revise process models based on the past logged executions in order to guide a majority of future workflow executions. This problem of revising process models using the execution log is call process model enhancement. Clearly, process model enhancement can benefit from techniques from process mining and process analysis. In this seminar, we plan to learn the basic problems and techniques of process discovery and model enhancement developed in the literature.
[You may need to be inside the UCSB domain to get access to publishers' links. If you are outside of the UCSB domain, here are a few options: Use a UCSB proxy server, the UCSB vpn, or just google and there is a good chance to find a link to the authors' copy. For more information on the first two, go to the UCSB library site --> help --> Off-Campus Access]
Tuesday, Jan 20, 3:00pm-4:00pm (one time change due to Jan 19 Holiday)
Paper to discuss: Repairing process models to reflect reality (authors: Dirk Fahland and Wil M. P. van der Aalst, Proceedings of BPM, 2012)
Presenter: Yan Tang
Monday, Jan 26
Monday, Feb 2
Monday, Feb 9
Monday, Feb 23
Monday, Mar 2
Monday, Mar 9