An application of data mining for marketing in telecommunication

Abstract

Summary form only given as follows. Churn is the process of customer turnover. In mobile telecommunications market, several techniques can be employed to analyze why customers churn and which customers are most likely to churn in the future. Many mobile telecommunications firms have a mobile agency that provides handsets installation, maintenance, and replacement support for their customers. Although most of these have some salesmen to handle day-to-day maintenance and small-scale troubleshooting, expert advice is often required from the manufacturing companies for more complex maintenance and repair jobs. Prompt response to a request is needed to maintain customer satisfaction. Therefore, a mobile agency is usually set up to answer frequently encountered problems from the customers. Such information can be utilized by marketing departments to better target recruitment campaigns and by active monitoring of the customer call base to highlight customers who may, by the signature in their usage pattern, be thinking of migrating to another provider. As a collaborative research project with a multi-national company, this research investigated the application of data mining techniques to extract knowledge from the customer service database for two kinds of customer service activities: decision support and customer's complaint analysis. The information stored in the customer service database are classified as structured and unstructured textual data. The structured data are mined to enhance the decision making process for better management of resources and marketing of products. The unstructured data are mined to enhance the decision making process after it is converted to structured data format. In order to mine the structured data in the customer service database, a data mining process based on the data mining tool, Clementine was proposed. To support customer's complaint analysis, a data mining technique based on answer tree and neural network. This data mining technique can operate within a system to provide efficient online customer's complaint analysis over the Internet (and intranet).