Accounting Logic Sequencing in SAP ECC: A Graph-to-Event-Log Transformation for Audit-Ready Process Mining

This paper addresses a core challenge in accounting information systems: reconstructing audit-compliant business process sequences from ERP accounting data that is structurally graph-based rather than sequentially ordered. Traditional timestamp-driven process mining fails to capture the logical order imposed by open item accounting, resulting in misleading process models for auditors. We propose a novel preprocessing algorithm that transforms financial process mining (FPM) instance graphs extracted from SAP ECC 6.0 accounting tables (BKPF/BSEG) into sequential event logs that respect accounting logic. Using a real-world dataset from a news agency's 2011 fiscal year, we demonstrate that our method reduces process model complexity by 53% and improves alignment with expected accounting workflows such as goods receipt → invoice → payment. This research enables auditors to leverage process mining directly on accounting data, bridging the gap between ERP financial records and actionable audit intelligence.