Semantic Cache Strategies For Continuous Nearest Neighbor Query And Mobile Learning

Nowadays mobile device is used at anytime and anywhere, to support Location Based Service (LBS), and to retrieve information such as nearest bank, nearest hotels, nearest restaurants, learning resource, from a remote server. The need to access such information is sometime important, urgent and unavoidable, for the mobile user.The usage of mobile devices to retrieve such information, from a remote server encountered some problems due to mobile user’s moving path, and due to the presence of an impermanent network environment, and weak connection. Therefore a mobile user will take a lot of time to retrieve information from a remote server, which makes his querying a time consuming process. Also Mobile Database limitations, such as, limited wireless bandwidth and the speed of communication, limited power (life of the battery), limited storage capacity, make difficult the use of mobile devices as an information retrieval device.Facing these problems it is strongly recommended to minimize the number of connections and volume of data transmission from a remote server. Data caching strategies will be a profitable solution for these problems.In order to improve LBS query,we proposed a Semantic Caching Strategy for Continuous Neighbor Query (CNNQ), in which a mobile user issues a query to retrieve the Nearest-Neighbor (NN) object such as (NN hotels, restaurants, school), of every point on his moving path. We used Voronoi index algorithm in server side to support CNNQ, which reduced the search space, by dividing the search space in a Voronoi Cell. Then in order to facilitate the used of mobile device as learning platform, we presented a Semantic Caching Strategy for Mobile Learning System. In this Strategy, a mobile user cached learning resource in a local cache according to his moving path.By storing learning resources in a local cache, a mobile user can quickly retrieve learning resources, which reduces his learning time and reduces the download time of resources.A simulation is conducted to examine the performance of our proposed Semantic Caching Strategy, in comparison with normal data caching and with the case where no cache is used. In our simulation to answer a CNNQ, we focus on a mobile user’s moving path with eleven NN objects; and presented two types of mobile learning resource: resource in pdf format (file.pdf) and resource image (file.jpg).The result shows that our proposed caching strategy gives a considerable improvement in response time and cache hit ratio, in comparison with normal data caching and with the case where no cache is used.