Cover Image

Telecommunication Subscription Fraud Detection Using Neural Network

Telecommunications fraud costs carriers billions of revenue dollars annually. The telecom fraud criminal is often ingenious and is typically motivated by money, anonymity, or both. Through use of the Internet, the fraud community is becoming more collaborative, and as a result, more ingenious. Beca...

Full description

Saved in:
Bibliographic Details
Main Author: Nuratikhah, Shahrin
Format: Thesis
Language:English
English
Published: 2008
Subjects:
TK5101-6720 Telecommunication
Online Access:http://etd.uum.edu.my/938/1/Nuratikhah_Shahrin.pdf
http://etd.uum.edu.my/938/2/Nuratikhah_Shahrin.pdf
http://etd.uum.edu.my/938/
Tags: Add Tag
No Tags, Be the first to tag this record!
id my.uum.etd.938
record_format eprints
spelling my.uum.etd.9382013-07-24T12:09:43Z http://etd.uum.edu.my/938/ Telecommunication Subscription Fraud Detection Using Neural Network Nuratikhah, Shahrin TK5101-6720 Telecommunication Telecommunications fraud costs carriers billions of revenue dollars annually. The telecom fraud criminal is often ingenious and is typically motivated by money, anonymity, or both. Through use of the Internet, the fraud community is becoming more collaborative, and as a result, more ingenious. Because of this, fraud detection applications must become more sophisticated to keep pace with the criminals. A system to prevent subscription fraud in fixed telecommunications with high impact on long-distance carriers is proposed. The system consists of a neural network technique, using Multilayer Perceptron (MLP). A total of 158 data samples from Telecom Malaysia Bhd were collected trained and tested using model. The prediction module allows identifying potential fraudulent customers at the time of subscription. The analysis of the data shows a reasonably strong correlation between the input variable, which consist of severity, user certainty factor, indicator case and target. The result shows that 78% of prediction accuracy has been obtained. From the result that has been produce, neural network has a potential to be used for detecting fraud in telecommunication. 2008-06 Thesis NonPeerReviewed application/pdf en http://etd.uum.edu.my/938/1/Nuratikhah_Shahrin.pdf application/pdf en http://etd.uum.edu.my/938/2/Nuratikhah_Shahrin.pdf Nuratikhah, Shahrin (2008) Telecommunication Subscription Fraud Detection Using Neural Network. Masters thesis, Universiti Utara Malaysia.
institution Universiti Utara Malaysia
building UUM Library
collection Institutional Repository
continent Asia
country Malaysia
content_provider Universiti Utara Malaysia
content_source UUM Electronic Theses
url_provider http://etd.uum.edu.my/
language English
English
topic TK5101-6720 Telecommunication
spellingShingle TK5101-6720 Telecommunication
Nuratikhah, Shahrin
Telecommunication Subscription Fraud Detection Using Neural Network
description Telecommunications fraud costs carriers billions of revenue dollars annually. The telecom fraud criminal is often ingenious and is typically motivated by money, anonymity, or both. Through use of the Internet, the fraud community is becoming more collaborative, and as a result, more ingenious. Because of this, fraud detection applications must become more sophisticated to keep pace with the criminals. A system to prevent subscription fraud in fixed telecommunications with high impact on long-distance carriers is proposed. The system consists of a neural network technique, using Multilayer Perceptron (MLP). A total of 158 data samples from Telecom Malaysia Bhd were collected trained and tested using model. The prediction module allows identifying potential fraudulent customers at the time of subscription. The analysis of the data shows a reasonably strong correlation between the input variable, which consist of severity, user certainty factor, indicator case and target. The result shows that 78% of prediction accuracy has been obtained. From the result that has been produce, neural network has a potential to be used for detecting fraud in telecommunication.
format Thesis
author Nuratikhah, Shahrin
author_facet Nuratikhah, Shahrin
author_sort Nuratikhah, Shahrin
title Telecommunication Subscription Fraud Detection Using Neural Network
title_short Telecommunication Subscription Fraud Detection Using Neural Network
title_full Telecommunication Subscription Fraud Detection Using Neural Network
title_fullStr Telecommunication Subscription Fraud Detection Using Neural Network
title_full_unstemmed Telecommunication Subscription Fraud Detection Using Neural Network
title_sort telecommunication subscription fraud detection using neural network
publishDate 2008
url http://etd.uum.edu.my/938/1/Nuratikhah_Shahrin.pdf
http://etd.uum.edu.my/938/2/Nuratikhah_Shahrin.pdf
http://etd.uum.edu.my/938/
_version_ 1644276307403472896
score 13.211869

Similar Items