Abstract:
In terms of all kinds of money transaction there occurs a common problem which is fraud. Fraud
is an increasing concept that can badly affect to the economy over the whole world. There are
various types of fraud such as mobile payment fraud, credit card fraud, bank fraud, insurance
fraud and other financial frauds. So, it is important to prevent fraud from financial transaction to
save the economy system. And predicting fraud is one of the efficient ways for preventing fraud.
The objective of our research is to detect fraud in mobile money transaction using data mining.
Actually, we will analysis a synthetic transactional dataset using classification and will predict
the probability of fraud in future transaction.
Description:
Mobile money transaction has become a very popular form of money transaction nowadays.
People all over the world use mobile payment system to make various type of payments. People
of rural areas are also now using mobile payment system for various purpose such as, from buying
product, transfer money to another account, cash out, to make recharge of others SIM etc. But
the matter of concern in these terms is fraud transaction. Fraud in those transactions can affect to
the economic condition of the normal people and also the whole country. Which can also affect
the economic condition of the whole world. So, to prevent fraud in transaction fraud detection is
an easy way. And we are doing this research to detect fraud in mobile money transaction using
data mining. As the real transactional data is unavailable due to the matter of privacy and
confidentiality, so we are using a synthetic dataset for doing our research. We have divided our
dataset into two parts such as training set and test set and applied simulation to determine the
accuracy of the test set. If the accuracy of the test set is fair enough then we can predict the fraud
in the next transactions which can be very useful to prevent fraud in mobile money transaction.
