The purpose of this paper is to propose a methodology to estimate the number of records that were omitted from a data set, and to assess its effectiveness.
The procedure to estimate the number of records that were omitted from a data set is based on Benford’s law. Empirical experiments are performed to illustrate the application of the procedure. In detail, two simulated Benford-conforming data sets are distorted and the procedure is then used to recover the original patterns of the data sets.
The effectiveness of the procedure seems to increase with the degree of conformity of the original data set with Benford’s law.
This work can be useful in auditing and economic crime detection, namely in identifying tax evasion.
This work is the first to propose Benford’s law as a tool to detect data evasion.
