User quality of experience estimation using social network analysis

With the ever-evolving number of new types of services alongside new paradigms like IoT (Internet of Things) and Pervasive Computing, finding the most appropriate service depending on users’ needs is difficult. Service providers try to deliver their service to the right user at the right time to have a higher conversion rate and customer loyalty, which is both dependent on and resulting in user satisfaction. For this reason, predicting the quality of experience (QoE) before providing the service and optimizing it for the user is one of the essential issues in the service-providing context. This paper addresses QoE prediction and proposes a new methodology for estimating the QoE; based on social network analysis. A new weighted distance metric is presented in this research based on community detection. This paper shows that using a novel link prediction method with community detection minimizes the error rate of QoE estimation. This novel method was tested on a video rating dataset and showed lower Mean Squared Error and higher accuracy compared to other research results. The paper’s main contribution is a QoE estimation framework that could be applied to any service type with measurable parameters.