Read out

Guest Talk: On the dimensions of predictive process monitoring

Dr. Chiara Di Fran­ces­co­ma­rino 

Date/Time: 11.04.2018, 15:00 

Loca­tion: D2.3.103 

Abstract 

Predic­tive process moni­to­ring aims at predic­ting the future of an ongoing process execu­tion by learning from past histo­rical busi­ness process execu­tions. Diffe­rent approa­ches have been proposed in the lite­ra­ture in order to provide predic­tions on the outcome, the remai­ning time, the required resources as well as the remai­ning activi­ties of an ongoing execu­tion, by lever­aging infor­ma­tion related to the control flow, the data flow, or even unstruc­tured text contained in event logs, reco­ding infor­ma­tion about process execu­tions. The approa­ches can be of diffe­rent nature and, some of them also equipped to offer users support in tasks such as para­meter selec­tion. This talk will provide an over­view of some of the approa­ches and the rese­arch direc­tions charac­te­ri­zing the field of predic­tive process moni­to­ring.

Bio 

Chiara Di Fran­ces­co­ma­rino is a rese­ar­cher at Fonda­zione Bruno Kessler (FBK) in the Process and Data Intel­li­gence (PDI) Unit. She received her PhD in Infor­ma­tion and Commu­ni­ca­tion Tech­no­lo­gies at FBK and Univer­sity of Trento, Italy, in 2011 by mainly inves­ti­ga­ting topics related to the semantic anno­ta­tion of busi­ness processes and the auto­mated reaso­ning capa­bi­li­ties that semantic anno­ta­tions can provide. She is currently working in the field of busi­ness process manage­ment, by exten­ding her rese­arch inte­rest from busi­ness process mode­ling to process execu­tion. Speci­fi­cally, she is currently inves­ti­ga­ting problems related to process moni­to­ring (e.g., in situa­tions of incom­plete execu­tion traces), process repair (given execu­tion traces non-­com­p­liant with the model), disco­very of hybrid (formal and informal) process models, as well as predic­tive process moni­to­ring based on histo­rical execu­tion traces.



Back to overview