Predictive Data Stream Filtering

In multimedia stream matching applications, user experience is becoming more and more important. In this paper, we study the problem of predictive stream matching, where a large number of queries are evaluated before the actual stream data arrive. This is equivalent to developing predictive algorithms for the filters registered on data streams. To this end, We propose to efficiently order the pipeline filters in multimedia streams under the Marko assumption. Our experimental evaluations validate that our predictive data stream filtering framework is able to provide an efficient solution for evaluating large number of queries on streams.