Conformidade em processos de negócio baseada em classificação de documentos e mineração de log de eventos

Abstract

The growing necessity for quality involves information systems and business processes. As a consequence, there is a growing demand to align the mapped and understood process of organizations with the performance of how the process is actually performed on a daily basis. With automated systems, the technology industry currently records events in information systems and generates data with ease, which are produced to generate value and insights to improve performance in the most diverse areas of organizations. In this work, we propose a study on conformance checking for generated models from the event log of an information system. A study that considers the main algorithms and metrics used in the literature to measure conformance checking in process mining, how the algorithms are defined and in which tools the approaches are tested. The proposed approach evaluates the semi-structured event log; for thus, text classification techniques are used to prepare the required structure of the event log. The main objective is to evaluate the conformance checking applied to the process mining area in order to analyze the extracted log, contextualizing the value approach with the definition of the existing process mapped from the manager understanding. To support this use in other data sets, the proposed model intends to be extensive for modification and use in other scenarios.