International Journal of Computer Science Issues
The problem in data mining applications is the mining of frequent patterns. Though there has been various techniques, such as pattern discovery, association rule mining etc, these methods generates a large volume of frequent patterns and rules which are not useful for finding the essential patterns among them, from the database. Alternatively, CFIM technique produces a relatively lesser number of closed frequent patterns. Patterns are pruned before clustering and the clustered patterns help in predicting the customer purchasing behavior which in turn helps to maintain an inventory and focus the point of sale on transaction data, enhancing sales.