Cache replacement strategy for streaming media based on semantic mining

In order to deal with the complicated network video data, the work focuses on the technologies of multilayer semantic mining and proxy cache for streaming media. Semantic mining method based on network parameter is presented. Quadratic regression popularity prediction model is introduced to analyze current popularity of streaming media. And then cache replacement strategy based on semantic mining is designed. Simulation results show that the cache replacement strategy is feasible and valid.