From Event Logs to Managerial Decision Intelligence: A Business Data Analytics Framework for Automated Workflow Discovery
Main article
Abstract
Event logs generated by enterprise systems provide a detailed empirical record of how business workflows actually unfold, yet many organizations still translate this evidence into managerial decisions through fragmented dashboards, manually selected process-mining settings, and ad hoc interpretations of discovered models. This study develops a business data analytics framework that connects event-log preparation, automated workflow discovery, multi-criteria model evaluation, and managerial decision intelligence. The framework is inspired by recent advances in multi-objective process discovery, but it reframes the technical discovery problem as a broader management analytics task: selecting a process model not only for statistical quality, but also for interpretability, governance relevance, and operational decision value. An illustrative analysis using four representative enterprise event logs demonstrates how automated configuration search improves the balance among fitness, precision, generalization, simplicity, and managerial usefulness when compared with default and expert-manual discovery strategies. The results indicate that the automated decision-intelligence setting raises the average decision usefulness score from 0.713 to 0.858 while preserving acceptable runtime performance through parallel evaluation. The study contributes to business data analytics by offering a structured bridge from process evidence to management action, showing how Pareto-based discovery and preference-aware model selection support bottleneck identification, compliance monitoring, resource allocation, and workflow redesign. Practical implications are provided for managers seeking to convert operational event data into auditable, repeatable, and strategically meaningful workflow intelligence.
