Article ID Journal Published Year Pages File Type
379883 Electronic Commerce Research and Applications 2010 12 Pages PDF
Abstract

We use customer clustering to explore the behavioral patterns of customers who subscribe to mobile services. Two clustering techniques, K-means and KVQ, are used to cluster customers using knowledge about attributes that are broadly grouped under usage, revenue, services, and user categories. We used inter-cluster analysis on the clusters generated from the two techniques to compare the distribution of customers among the different categories of attributes. We observed that it was important to use multiple techniques for clustering. Our analysis discovered several interesting facts about customers, such as the imbalance between customers’ usage of mobile services, subscriptions to services, and revenue contributions. These knowledge nuggets could enable mobile service providers to better align their marketing strategies with the needs of customers.

Related Topics
Physical Sciences and Engineering Computer Science Artificial Intelligence
Authors
, ,