Masters thesis, Universidade de Brasília, Brasília.
ISSN/ISBN: Not available at this time. DOI: Not available at this time.
Note - this is a foreign language paper: POR
Abstract: With the evolution of science and technology, new ways to detect accounting fraud were conceived. The Newcomb-Benford Law (LNB),or simply Benford's Law, proves to be a simple and effective tool and can be adapted to identify accounting fraud by comparing the frequency of the first digits against a standard established empirically by Benford. Using artificial intelligence methodologies and machine learning mechanisms adaptive tools can be developed, for different types of fraud. This study seeks to validate the application of Benford'sLaw through artificial neural networks and to provide subsidies for the work of tax auditors, so that it can contribute not only to reducing fraud and greater agility in its detection but also, to increase the reliability and transparency of the data made available to the market and society, also providing greater reliability to economic analyzes. This work developed a model that makes statistical analysis of the data provided by the Superior Electoral Court and used the model to analyze the data from the last elections in Brazil. The research findings generally suggest that the distributions found in 2016 and 2020 follow Benford's Law, while in 2014 and 2018, the analyzes suggest non-conformities.
Bibtex:
@mastersThesis{,
AUTHOR = {Ramos, Paulo César Roxo},
TITLE = {Lei de Benford : uma Integração no Trabalho de Auditoria},
SCHOOL = {Universidade de Brasília},
YEAR = {2021},
ADDRESS ={Brasilia},
}
Reference Type: Thesis
Subject Area(s): Accounting, Computer Science, Voting Fraud