Research on caching strategy in a VoiceXML-based Mobile Learning System

In order to push the new voice-based mobile learning service more quickly and better and decrease the operating risk, employing the technology of the voice extensible markup language (VoiceXML) as the service logic language, an independent VoiceXML-based mobile learning system (VMLS) is built, its hardware and software system structure design is achieved. The system will be open to course content providers and enable them to provide more new services to students. In the meantime, aiming to the network delay, the key problem existing in VMLS, an adaptive Markov prefetching algorithm shared by multi-users and a more efficient caching replacement strategy are provided, which can improve the forecasting accuracy rate and the overall performance of cache system. The research results are not only used in VMLS, and are also useful the other voice-based application system and the other related research fields.