BENFORD'S LAW AND ARTIFICIAL INTELLIGENCE

AN INTEGRATION IN AUDIT WORK FROM A REGIONAL PERSPECTIVE

Authors

  • Paulo César Roxo Ramos Mestrado em Economia: ECO/FACE/UnB
  • Roberto de Goes Ellery Junior Professor da ECO/FACE/UnB
  • Antônio Nascimento Junior Professor ADM/FACE/UnB, anjunior@unb.br

Keywords:

Benford; Inteligência Artificial; Detecção de Fraude; Eleições; Processos Naturais

Abstract

Abstract

The evolution of science and technology has brought new ways to detect accounting fraud. The Newcomb-Benford Law (LNB) is a simple and effective tool and can be adopted to identify accounting frauds by comparing the frequency of the first digit against a standard empirically pattern established by Benford. The use of Artificial Intelligence (AI) methodologies and machine learning method allows the development of adaptive tools for different types of fraud. This work presents a model for analyzing data provided by the Superior Electoral Court and employs the model for analyzing data from the latest updates in Brazil. The research finds that distributions found in 2016 and 2020 follow Benford's Law, while in 2014 and 2018may not.

Author Biographies

Paulo César Roxo Ramos , Mestrado em Economia: ECO/FACE/UnB

Mestrado em Economia: ECO/FACE/UnB

Roberto de Goes Ellery Junior , Professor da ECO/FACE/UnB

Doutor em Economia. Professor da ECO/FACE/UnB

Antônio Nascimento Junior, Professor ADM/FACE/UnB, anjunior@unb.br

Doutor em Economia Professor ADM/FACE/UnB, anjunior@unb.br

Published

2021-12-19