IPTV user QoE prediction based on the LSTM network

Abstract

Listen

Translate article

With the rapid development of Internet Protocol Television (IPTV), operators are more interested in user Quality of Experience (QoE). The aim of QoE prediction is to study the relationship between the user QoE and the features selected from IPTV set-top box. To handle this issue, in this paper, we propose an IPTV user QoE prediction algorithm based on the long short-term memory (LSTM) network. Specifically, we firstly collect and analyze the dataset, selecting some important features both from subjective and objective aspects. Then we use the LSTM network to perform QoE modeling. Finally, we compare the performance of the proposed algorithm to those of k-nearest neighbor (KNN), support vector machine (SVM) and conventional neural network. Experimental results show that the proposed algorithm presents high performance in QoE prediction.