Seleção de atributos para mineração de processos na gestão de incidentes

TLDR: The solution presented in this paper represents an improvement in process mining on the specific context of creation annotated transition system and its use as a statistics generator for the whole modeled process.

Abstract: AMARAL, Claudio Aparecido Lira do. “Attribute selection for process mining on incident management process”. 2018. 136 p. Dissertation (Master of Science) – School of Arts, Sciences and Humanities, University of São Paulo, São Paulo, 2018. The incident management process is the most widely adopted by companies. However, still lacks techniques that can generate precise estimates for the completion time. This work performs a study in a real incident management process, by means of process mining, able to find out the real process model in the form of annotated transition system and propose automated means for selecting attributes that describe it accordingly, in order to generate realistic estimates of the time to conclusion. The resulting strategy of application feature selection techniques filter and wrapper is able to provide generation of more accurate annotated transition systems with some degree of generalization. The solution presented in this paper represents an improvement in process mining on the specific context of creation annotated transition system and its use as a statistics generator for the whole modeled process.