Read out

Guest Talk: Checking Process Compliance on the Basis of Uncertain Event-to-Activity Mappings

Henrik Leopold, PhD 

Date/Time: 12.07.2017, 12:00 

Loca­tion: D2.2.094 


A crucial requi­re­ment for comp­li­an­ce-che­cking tech­ni­ques is that observed beha­vior, captured in event traces, can be mapped to the process models that specify allowed beha­vior. Without a mapping, it is not possible to deter­mine if observed beha­vior is comp­liant or not. A considerable problem in this regard is that esta­blis­hing a mapping between events and process model activi­ties is an inher­ently uncer­tain task. Since the use of a parti­cular mapping directly influ­ences the comp­li­ance of a trace to a speci­fi­ca­tion, this uncer­tainty repres­ents a major issue for comp­li­ance checking. To over­come this issue, we intro­duce a proba­bi­listic comp­li­an­ce-che­cking method that can deal with uncer­tain mappings. Our method avoids the need to select a single mapping, but rather works on a spec­trum of possible mappings. A quan­ti­ta­tive evalua­tion demons­trates that our method can be applied on a considerable number of real-world processes where tradi­tional comp­li­an­ce-che­cking methods fail.


Henrik Leopold is an Assis­tant Professor at the Depart­ment of Computer Science at the VU Univer­sity Amsterdam. He received a PhD degree (Dr. rer. pol.) as well as a master degree in infor­ma­tion systems from the Humboldt Univer­sity Berlin. His doctoral thesis on Natural Language in Busi­ness Process Models received the TARGION Disser­ta­tion Award for the best doctoral thesis in the field of Infor­ma­tion Manage­ment between 2012 and 2014. From July 2013 to March 2014 he worked as a post­doc­toral rese­arch fellow at the Humboldt Univer­sity Berlin. After­wards, he joined WU Vienna as an Assis­tant Professor from April 2014 to January 2015. He has been a visiting rese­ar­cher at the Eind­hoven Univer­sity of Tech­no­logy, the Univer­si­dade Federal do Estado do Rio de Janeiro, and the Univer­sity of Mann­heim. His current rese­arch inte­rests relate to the combi­na­tion of Busi­ness Process Manage­ment with Natural Language Proces­sing and Arti­fi­cial Intel­li­gence tech­ni­ques. The results of his rese­arch have been published, among others, in Data & Know­ledge Engi­nee­ring, Deci­sion Support Systems, IEEE Soft­ware, IEEE Tran­sac­tions on Soft­ware Engi­nee­ring, the Journal of Systems and Soft­ware, and Infor­ma­tion Systems.

Back to overview