Maximizing customer satisfaction in maintenance of software product family

Customer satisfaction and cost are very important factors in software maintenance. However, the tradeoff between them is formidable in maintenance of software_product family. This paper suggests a decision tree method to improve multi-customer satisfaction adaptively with the available human resource. An experiment was conducted in the maintenance department of a Chinese software vendor who has provided more than 1000 hotels with its proprietary hotel information management system. The machine learning approach improved customer satisfaction with the same human resource cost