Scalable attack analysis of business process based on decision mining classification

Banking crime is one of the widespread phenomena in 2016 are closely associated with the used of computer-based technology and internet networks that constantly evolving. One of them is the burglary of customer accounts through the internet banking facility. To overcome this, we need a method of how to detect a conspiracy of bank burglary case of customer accounts. The way to scalable is by get a mining decision to get a decision tree and from the decision tree to get a decision attribute value to determine the level of anomalies. Then of all the attributes decision point is calculated rate of fraud. The rate of fraud is classified through level of security of attack by the attacker then entropy gain is used to calculate the relative effort between the level of attacks in the decision tree. The results show that the method could classify three levels of attacks and the corresponding entropy gains. The paper uses decision trees algorithm, alpha++ and dotted chart analysis to analyze an attack that can be scalable. The results of the analysis show that the accuracy achieved by 0.87%.