Fraud detection

Fraud detection was addressed in the preceding chapter (Chapter 6, Node Importance). We talked about the way fraudsters create crime rings and collaborate with each other to avoid detection. This kind of organization is hard to detect by traditional methods, but graphs, thanks to their native structure based on relationships, are a very helpful tool to fight against all types of fraud. Assuming that fraudsters will interact more with each other, like in the case of fake users sharing the same phone number or address, graphs can form a community and be more easily identified.