Fear Not the Software
Most data analysis for fraud is specifically focused on precise red flags of fraud. The tools used are therefore query-based, which allows the fraud examiner to extract records that meet a certain criteria such as invoices paid to a vendor who has the same address as an employee. The process is inductive in nature and while it'll find much of the fraud it also misses the uncommon schemes that could sink an organization. Let's face it - any good fraudster knows what's commonly searched for and therefore will develop schemes that are missed by such detection reports.
Here we'll present deductive and holistic data analysis techniques to find the uncommon frauds that are less predictable to the fraudster. These techniques work like divining rods that don't necessarily provide an exact answer ("there are five employees that are set up to receive vendor payments") but rather point us in the general right direction ("there are 23 invoices that make up 65 percent of the the payment value made during the year"). Because you may be trained to look at specifics these three directions could be new for you:
1. Statistical view
2. Auto-rule generator view
3. Data visualization view
We'll use practical examples and free demo copies of the associated software products.
For full access to story, members may sign in here.
Not a member? Click here to Join Now.
Or Click here to sign up for a FREE TRIAL.