Digital Library

cab1

 
Title:      FRAUD DETECTION IN ELECTRONIC TRANSACTIONS
Author(s):      José Felipe Jr., Adriano Veloso, Wagner Meira Jr., Adriano Pereira
ISBN:      978-989-8533-01-2
Editors:      Bebo White, Pedro Isaías and Flávia Maria Santoro
Year:      2011
Edition:      Single
Keywords:      Fraud Detection, E-Commerce, Web Application, Data Mining
Type:      Full Paper
First Page:      333
Last Page:      339
Language:      English
Cover:      cover          
Full Contents:      click to dowload Download
Paper Abstract:      Due to the significant increase of fraud cases resulting in billions of dollars losses each year worldwide, the study and development of techniques that assist in their identification is essential to deal with this problem. Data mining has played a strong role in this task and can be used to identify in advance transactions and behaviors that indicate fraud. This work aims to apply and evaluate a Lazy Associative Classification (LAC) technique to identify fraud in electronic transactions conducted by the most popular Brazilian electronic payment service. Our results show that the LAC algorithm has a good performance in fraud detection, where it was possible to cover 100% of the chargebacks with just 5% of the dataset. This result is promising since it was obtained from a dataset that presents a skewed distribution of the fraud classes. Moreover, we perform an economic efficiency analysis of the results in order to understand the real scenario from the company point of view.
   

Social Media Links

Search

Login