Sampling and abstract interpretation for tackling noise in process discovery
Tutor / director / evaluatorCarmona Vargas, Josep
Document typeMaster thesis
Rights accessOpen Access
Nowadays, business processes rely more and more on the information systems, making them essential for an organization. From hospitals that record the histories of the patients to banks making thousands of transactions per day, information systems produce data that can be very valuable to their owners. Then, the challenge is to exploit event data in a meaningful way to be able to analyze the processes based on the information found. That is the objective of Process mining, a research discipline that provides a bridge between data mining and business process modeling (BPM). (...) At the moment, very few techniques in the literature address the presence of noise. The way these algorithms deal with noise and incompleteness is similar. The contribution of this thesis is a novel strategy to deal with noise that deviates from the methods that have been traditionally used to tackle noise.