Temporal Data Storage and Analysis in Research Information Systems: An Event-Sourced Bitemporal Architecture for KDSF and CERIF

Current Research Information Systems (CRIS) frequently suffer from anachronistic attribution bias and high query latency when calculating longitudinal scientometric indicators over evolving institutional hierarchies. This preprint introduces an Event-Sourced bitemporal architecture that decouples transaction ingestion from point-in-time analytical projections (CQRS). We formalize an institutional succession algebra for department mergers and splits, design a projection engine compliant with the German Kerndatensatz Forschung (KDSF 2.1) and euroCRIS CERIF standards, and validate performance over a synthetic 30-year institutional dataset. Furthermore, we address the immutability conflict with GDPR Article 17 via cryptographic erasure (crypto-shredding). Note: The complete text of this research is provided in both English and German within the attached PDF files.