In this paper, we proposed a method for mining mobile users’ Quality of Experience (OoE) model based on weighted LDA. In the recent years, QoE has become an important concept for the quality of networks and services. At present , QoE has attracted the interest of network operators and service providers, because of providing a good QoE service to their customers can satisfy the customers and bring more users. In this paper, we are trying to build up users’ QoE model through topic model, an approach to generate a generative model for data mining. Latent Dirichlet Allocation (LDA) is a feasible and effective algorithm in text modeling. We propose an weighted LDA-based interest model within the modeling framework, and evaluate it on a mobile network users’ behavior extraction system. In this system, we can analyze the users’ behaviors, and build up a vector model for each user through a simple way. Besides, with the help of the topic model, we can get an exact model for users’ QoE, because we can generate the topic model through the vector model. Thus we can get the users’ QoE model, through which we can learn each user’s experience. Then the network operators can provide a better network service for their customers. In the end, we elaborate QoE management requirements for mobile network scenarios, and provide a QoE modeling approach for the mobile network scenarios.
A Survey of Standardized Approaches towards the Quality of Experience Evaluation for Video Services: An ITU Perspective
