Object-oriented Bayesian networks for combining several features of the quality

Customer orientation can be a strategic tool to support management decisions. Private companies and public authorities carry out customer satisfaction surveys to measure the perceived quality. The quality is a dynamic feature often interpreted as a mix of satisfaction items that can be analyzed both separately and jointly. Moreover, they can contribute to produce an index that synthesizes the hidden global quality. Here, we propose to combine several aspects of satisfaction using Object-Oriented Bayesian Networks.We present an application where each satisfaction area is modelled by a Bayesian network learnt from data; an Object-Oriented Bayesian network is built to handle the domain as a whole. The tool can then be used to evaluate improvement actions developed in one or more areas.