DISCOVER MINER: A SIMPLE APPROACH TO OPTIMIZE PROCESS MINING MODELLING

Abstract

This article presents the Discover-Miner platform that apply four process mining models: Alpha Miner, Discover-Miner, Heuristic Miner, and X_Processes. The objective was to identify the best mining model, evaluating it based in one of three metrics: Fitness, Precision, and Generalization. The platform generates the metrics and Petri nets for each processed log. The methodology included the systematic application of the models to event logs, evaluating the results with the chosen metrics. It was concluded that the platform is useful in evaluating process mining models, helping managers choose the most suitable model. The study also provides guidelines for future research and practical applications, aiming to improve the efficiency and accuracy of business analyses.