SAS arms IRS with new fraud detection tools

The Internal Revenue Service is getting a new weapon in its ceaseless battle to detect, prevent and resolve criminal and civil noncompliance with tax law.

SAS, a business analytics software and services company, has won a $6.25 million contract to support IRS’s new electronic Return Review Program system, which is designed to help reduce the $345 billion tax gap.

That’s the difference between what taxpayers owe and what they pay voluntarily and on time, a Dec. 8 SAS announcement explained.

SAS technology can improve fraud detection and uncover noncompliance when tax returns initially are filed, reducing the issuance of questionable refunds, it said.

The contract will give the IRS Return Review Program access to many SAS technologies, including SAS Fraud Framework for Government, SAS Social Network Analysis and SAS' data integration, data mining and business intelligence technologies.

To analyze massive amounts of data, the IRS will have big data analytics capabilities through SAS In-Database functionality, SAS Scoring Accelerator for Greenplum and SAS Grid Manager.

SAS scores tax returns through a hybrid approach of business rules, anomaly detection, predictive modeling and social network analysis. Users can set up business rules that detect possible fraud and immediately alert investigators or auditors to suspect returns.

The software searches data for anomalies that could indicate fraud or error. Predictive modeling uses historical behavioral information to identify suspicious behaviors similar to known fraud patterns, the announcement said.

Social network analysis uncovers hidden relationships or linkages that suggest collusion and organized fraud rings.

SAS Text Miner, another weapon against fraud, scours unstructured data, such as call center data, to detect suspicious activity. Alerts and results are reported via a customizable dashboard, and SAS' case management capabilities help investigators prioritize and assign cases.

SAS combines an array of technologies to tackle fraud, waste and improper payments in many areas of government – from detecting collusive patterns in entitlement programs such as Medicare and Medicaid to purchase-card fraud, bid-rigging and terrorist financing.

Fifteen federal agencies and 80 national governments use SAS software's analytic and predictive capabilities for critical initiatives such as performance management, budgeting, logistics, cyber defense, combating fraud and improper payments, and threat assessment, the company said.

SAS, which originally stood for “statistical analysis system," was founded in 1976 at North Carolina State University to analyze agricultural research.

In 2010 the Cary, N.C., company reported worldwide revenue of $2.43 billion.