Detecting fraud using modified benford analysis

Large enterprises frequently enforce accounting limits to reduce the impact of f raud. As a complement to accounting limits, auditors use Benford analysis to det ect traces of undesirable or illegal activities in accounting data. Unfortunatel y, the two fraud fighting measures often do not work well together. Accounting l imits may significantly disturb the digit distribution examined by Benford analy sis, leading to high false alarm rates, additional investigations and, ultimatel y, higher costs. To better handle accounting limits, this paper describes a modi fied Benford analysis technique where a cut-off log-normal distribution derived from the accounting limits and other properties of the data replaces the distrib ution used in Benford analysis. Experiments with simulated and real-world data d emonstrate that the modified Benford analysis technique significantly reduces fa lse positive errors.