Abstract: Firstly, based on the  cloud computing application model, we proposed a double\|layer cache technology which could efficiently utilize the cloud storage architecture. By establishing the distributed  cache between the client and the server, it could  effectively avoid users frequent access to remote data and build lightweight clients for users, which solved the problem that current geoscience data visualization software occupied a large number of user’s local storage capacity, and adapt to the rapid development of mobile devices. In the mean time, the server side  also avoided multiple access to the cloud storage file system, reducing a lot of data retrieval and loading time. Secondly, we  proposed an association rule last successor access prediction algorithm, according to user’s historical access records, the association rules were used to mine the user’s access mode, and predict their access behavior. Then the data was loaded in advance, the cache hit rate was improved, we solved  the problem of constantly moving region of interest and changing the rendering data frequently in the process of visualization, our system could effectively deal with the user’s multiple  access patterns case and improve the accuracy of the prediction. 
Experimental results show that the cloud storage architecture  significantly reduces the  local resource consumption. The accuracy rate of the  access prediction algorithm is 47.59% in the worst case, the average accuracy rate is 913%, and the average cache hit rate of distributed cache is  9561%, which can effectively support the rapid visualization of large-scale seismic data in the cloud.

Key words: cloud storage architecture, double-layer cache, large , data index, access prediction, rapid visualization, network communication