Date Added: Jul 2012
This paper proposes an enhanced design for development of computational model for intelligent selection of telecom services meant for telecom customers. Computational model comprises of Data mining and soft computing techniques. CLARA cluster algorithm as Data mining technique is used to categorize telecom customers into five clusters referring to five different telecom services scheme. Categorization of customers is based on their Historical call usage pattern. CLARA addresses the noise sensitivity defect of K-means which was used in previous model. Consequently soft computing technique such as neural network is implemented to train the system. Neural network performance graph shows that this model gives better result than the previous model.