Інструмент для аналізу та візуалізації бізнес-процесів на основі логів подій

Abstract

This paper presents a tool for analyzing and visualizing business processes based on event logs generated by information systems. The proposed approach aims to reconstruct the actual "as-is" process using event data recorded in CRM, ERP, Service Desk, workflow platforms, and other systems that generate event logs. The tool supports preprocessing of event logs, filtering noisy and incomplete records, building a weighted process graph, calculating performance metrics, identifying bottlenecks, and estimating process variability. Particular attention is given to comparing models obtained from the full dataset with those built from cleaned data, as well as to rebuilding the process model after applying filters by time period, case type, performer, or selected subsets of cases. The expected result is an interpretable process model that helps identify delays, frequent execution paths, deviations, and opportunities for further process improvement.