Data Mining Used by New York to Catch Tax Fraud

A new data mining system, similar to the ones used by Homeland Security agencies, is being used by the New York State Department of Taxation to catch tax fraud by individuals and businesses.

The use of data mining to catch tax cheats is the brainchild of Bill Comiskey, deputy commissioner fo the Office of Tax Enforcement in New York.

Comiskey has spent most of his professional life investigating financial fraud and he collected a record $3 billion in tax revenue for New York during his first year as deputy tax commissioner ( $2 billion more than the year before he took the job ).

New York’s data mining system will pull personal and financial data on people and businesses from a broad range government and third party sources, including insurance company data, mortgage application data, wholesale invoices to businesses and other public and private records sources.

The department of taxation is even planning to use transaction records from credit card companies in their tax mining system.

This continuous flow of personal and business data will then be compared with reported income on tax filings and analyzed for warning signs of tax fraud.

Nowadays, nearly every bit of personal and business data is being mined by New York tax officials to go after the cash-based “underground economy” in the state.

The new data mining system is said to make New York trendsetter in the use of computer analytics to catch unreported income on tax filings.

Aiding Comiskey’s data mining efforts is the fact that the New York state legislature recently made it a crime to lie to tax fraud investigators.

Criminal tax fraud investigations in New York grew from 581 during the 2006 – 2007 fiscal year to 2078 in the 2008 – 2009 fiscal year.

In addition, staffing at New York State’s Department of Taxation has grown from 2,562 in 2006 to 3,060 currently and the department is hiring 111 new tax auditors.

The new data mining system is actually based on a data analysis tool that is used by IBM to design software that was then customized for the state of New York’s tax mining purposes.

Comiskey wants to see his tax data mining system contain as much personal and business information as possible, which can then by sorted and analyzed to raise red flags on possible tax fraud.

[ Source: ]