Association analysis and prediction for IPTV service data and user´s QoE

In order to provide good service and improve user´s feeling and degree of satisfaction, video service providers are now interested in understanding the influence of attributes on Quality of Experience (QoE). In this paper, based on IPTV business, we study the relationship between alarming data from IPTV set-top boxes and the user´s QoE. First, data cleaning and analysis are performed. After these procedures, the important attributes influencing QoE are selected. Then decision tree is used for modeling dataset. Here information entropy minimization heuristic is adopted to discretize continuous-valued attributes and the improved C4.5 algorithm is designed to fastly and accurately build the tree. Experimental results show that the proposed scheme can indeed improve the prediction accuracy when compared with other competing schemes.