Author : Ramadan Moawad
Source : Journal of Al Azhar University
Date of Publication : 01/2010
Abstract :
abstract - Customer relationship management “CRM” is very important factor in enhancing the organization competitiveness. In this paper, Data mining “DM” techniques are used to improve customer services in a radiology centers. Clustering customers is needed to find unsatisfied need, promote services packages and create new service packages. The proposed system radiology data mining system “RDMS” consists of three components; preprocessing, clustering and post processing. The data collected is for a period of four month for 6700 transaction. Three data sets are constructed from the original data set by dividing the whole data into 90%, 85% and 80% for training and 10%, 15% and 20% for testing respectively. Three K-means model are used with k=10, 15 and 18 cluster and each data set is used to calibrate and test the model for a total of nine ones. It is found that the best model is the one with 15 clusters. The clustering results are represented to a medical specialist who found that some results are reasonable and others go along with the center type and its policy.
Download PDF