Data Mining Application in Supporting Fraud Detection On Ethio-Mobile Services
No Thumbnail Available
Date
2006-03
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
The problem of Mobile frauds has been getting more and more serious for many years,
and is even getting more and more worse not only in western countries but also in some
developing countries. Fraud is the most significant threat to the communications
business, eroding margins, consuming network capacity and jeopardizing customer
relationships. Detection, Analysis and prevention mechanisms are emerging both from
telecommunications operators and academia. In this paper, the possible application of
data mining in supporting fraud detection on Ethio-Mobile Services has been tested by
the use of neural network technique.
The methodology used for this research had three basic steps. These were: data
collection, data preparation, and model building and testing. The required data was
collected from Ethiopian Telecommunication Corporation which is called Call Detail
Record. This record shows the behavior of each mobile phone users. Next, data
preparation tasks (such as data cleaning, feature selection, data transformation etc) were
undel1aken. Several Neural network models were built and tested for their classification
accuracy; and the model with encouraging results was taken.
Description
Keywords
Information Science