Multi-semantic popularity model of streamingmedia based on data packet analysis
基于数据包分析的流媒体多语义流行度模型

With the rapid development of streaming media technology and application, there are many pressures for network bandwidth and system resources, and it becomes more and more important for the semantic mining of network information and the analysis of interest preference. So this paper used extend uniform content locatorexUCL model to describe the semantic information in each layer of the network protocol. According to the data packet analysis of streaming media, it established the multi-semantic popularity model, and the same time it was verified with Chi-square method and real network data. This model has obvious effect in improving the efficiency of cache replacement algorithm of media resources, and the quality of customer service in streaming media system.