Privacy preserving in banking sector
Shashidhar Virupaksha, V. Bhramaramba · 2016
Banking is a financial institution that allows lending money to a borrower and it gives deposits to the customers based on bank balance sheet. Customer profiling consists of customer details, customer attractiveness, and satisfaction. This data can be obtained from bank account transactions, loan applications, and loan repayments. Therefore, it maintains huge amount of customer information, credit card usage pattern, transactions and so on. Some of the data mining techniques like clustering can be used for either forecasting or description and segmentation helps to understand customer resources based on information. Here an intruder can get the sensitive information of a particular person even if the removal of personal identifiable information like name, address etc. Intruders can get the information based on QI value in the dataset which in known as membership disclosure attack. K-anonymity doesn't provide appropriated privacy for the data against those attacks. While using data mining techniques there gets a privacy problem, in this paper it provides an effective privacy definition called l-diversity. In this there are `l' well represented values for sensitive attribute.