Designing a Recommender System for Optimizing and Managing Bank Facilities through the Utilization of Clustering and Classification Algorithms

Bank facilities, as one of the most important functions of banks, are very important to the survival of banks. The importance of identifying facility consumers is more than depositor customers. In previous studies, the importance of such customers has not been well considered. This study aims at designing recommender systems for predicting the customer behavior in receiving facilities, using clustering and classification algorithms. The designed system helps banks to identify types of facility demanding costumers, and to make decisions related to each customer cluster in order to make maximum profits and reduce the cost related to loans and receivables. The system accurately predicts each customer's cluster and also forecasts account balance of each customer group according to the current data set. Based on Analyses carried out on the clusters and related data sets of clients, is the recommender system is provided for to branch users. In order to design the recommender system, clustering and classification outputs were also deeply validated. The margin of error was very low and therefore, data mining outputs were verified and used to develop the final system as well as a user interface for the final utilization of recommender system which was presented and validated by the supervisors of bank. Supervisors also provided some suggestions for improving the final solution.