Anomaly Detection and Categorization for a Data Quality Management Framework in Financial Regulatory Reporting

Financial institutions are subject to stringent regulatory reporting requirements to manage operational risk in international financial markets. Producing accurate and timely reports has raised challenges in current data processes of big data heterogeneity, system interoperability and enterprise-wide management. Data quality management is a key concern, with current approaches being time-consuming, expensive, and risky. This research proposes to design, develop, and evaluate a Financial Reporting Data Quality Framework that allows non-IT data consumers to contextualize data observations. The framework will use anomaly algorithms to detect and categorize observations as genuine business activities or data quality issues. To ensure sustainability and ongoing relevance, the framework will also embed an update mechanism.