Using Advanced Analytics to Predict Risk for Grants Oversight
MetadataShow full item record
While there is much discussion about applying advanced analytic methods to the auditing and oversight fields, there has been little discussion in the academic literature about using these methods for oversight. The A-133 single audit data is a unique data set that can only be maximized using advanced analytic processes due to its size and current structure. This project applies text mining and predictive modeling techniques to this data set in order to determine both the feasibility and benefits of using these methods for grants management oversight. Using these methods, I was able to identify 12 percent more findings in the audit reports than I was able to identify using established, quantitative methods. This project establishes that advanced analytics methods can be a useful for supporting grant oversight and supporting agencies’ efforts to target resources on grant recipients who are highest risk for fraud, waste, and abuse.