A novel user satisfaction prediction model for future network provisioning

Abstract

The notion of user perception has grown in terms of its importance and complexity. This paper presents results of an experimental study focused on predictive modeling of the relations between the user perception, user satisfaction and objective technical parameters in data communication services. A new model for prediction of user satisfaction was devised using probability theory based on Markov chain. Two experiments were completed for web browsing scenarios. The results of the first experiment have confirmed that previous user experience has significant effect on the user perception of quality and should represent a vital element of future predictive user models. The result of the second experiment is a user satisfaction prediction model, which presents a novel insight and deeper understanding of user perception of quality. This model can significantly improve level of user satisfaction with services in telecommunications systems if implemented within advanced system design, optimization and quality assurance procedures.