Model heterogeneous fraud and contamination propagation

Model heterogeneous forms of fraud by distinguishing authenticity fraud targeted by the Data Credit Guarantee Organization from errors in data content, and derive how contamination propagates to secondary data products.

Background

The simulation restricts fraud to a single type: data that cause a loss to the buyer. The paper notes that the Data Credit Guarantee Organization primarily targets authenticity-related problems, whereas errors in substantive data content may have different detection and impact dynamics. It also does not represent how contaminated or fraudulent data propagate when they are incorporated into derivative or secondary datasets. Extending the model to distinguish these forms of harm and to represent downstream contamination would broaden the analysis of trust and risk in data markets.