Crop Insurance Today August 2012 - (Page 23)

TODAYcrop insurance RMA/AIP Data Mining Steering Committee By Troy Brady and David Hall, NCIS The Agricultural Risk Protection Act of 2000 (ARPA) required the Secretary of Agriculture to use data technologies to administer and enforce crop insurance program compliance and integrity initiatives. The Risk Management Agency’s (RMA) Office of Strategic Data Acquisition and Analysis (SDAA) was established to manage the Agency’s data warehousing and data mining program. SDAA partnered with the Center for Agribusiness Excellence (CAE) at Tarleton State University in Stephenville, Texas to carry out its responsibilities. Since 2001, CAE has developed and managed RMA’s data warehouse and data mining initiatives. As RMA’s data mining function has grown over the years, Approved Insurance Providers (AIPs) have begun to take a more active role in the data mining program. monitoring program required by ARPA. These are producers whose data are indicative of anomalous behavior. Similarly, CAE provides AIPs with a spot check list of anomalous producers, agents and loss adjusters each year. CAE has developed a relational database that contains records from multiple RMA and FSA data tables (Figure 1). It includes all Federal Crop Insurance Corporation (FCIC) reinsurance year policyholder data from 1991 through the present. The data warehouse and data mining techniques are used to identify agents, loss adjusters and producers that exhibit “anomalous” claim outcomes. Anomalous outcomes are defined as outcomes that are equal to or greater than 150 percent of the mean claim outcome in a designated area. Although anomalous behavior is not considered evidence of fraud, waste and abuse, it is an indicator of situations that may warrant further examination. Extensive Data Warehouse RMA and FSA program data are only a portion of the information within RMA’s Figure 1. The Data Mining Process Input Agency data Other federal data Data from public sources Analysis Output Background on Current Data Mining Structure and Activities Data mining assists RMA in identification and elimination of program fraud, waste and abuse by targeting program weaknesses and identifying agents, loss adjusters and producers that may need to be monitored or investigated. CAE is also responsible for delivering an annual spot check list of producers to the Farm Service Agency (FSA) as part of their Iterative query Collected, linked, validated,and formatted Data warehouse Pattern-based query Results Results Results Results Subject-based query Iterative query Source: GAO, adapted from Vipin Kumar and Mohammed J. Zaki. Data from the private sector Analysis can be iterative with the Results can be in results of one query being used to printed or electronic define criteria for a subsequent query format CROP INSURANCE TODAY 23

Table of Contents for the Digital Edition of Crop Insurance Today August 2012

“What is, and What Should Never Be....”
Developing Risk Management Plans
Crop Insurance Rate of Return: Issues & Concerns
2011 Research Review
RMA/AIP Data Mining Steering Committee
NCIS Adjuster Training in Full Swing
Industry Support of FFA
Crop Insurance & Specialty Crops

Crop Insurance Today August 2012