GRID-BASED HYBRID DATA MINING DEVICE AND METHOD THEREOF

摘要

A grid-based hybrid data mining device and a method thereof are provided to increase the usability of data/grid by using geographically distributed computing resources, and shorten a processing time by dynamically distributing tasks according to computing the capability of the grid resources and reducing probable bottleneck generated in huge task processing/execution. A task manager(10) includes a task requesting module requesting the resources of grid computing for a huge data mining task, a task managing module distributing the huge data mining task to a task distributor, and a result testing module testing a data pattern recognition rate of the processed data. A grid resource intermediary(30) includes a plurality of resource intermediating module dynamically requesting the resources by receiving distribution of the task managing module and transmitting a result produced by integrating the processed data to the result testing module. A plurality of resource providers include a data mining modules producing a learning result and a test result for data mining by receiving a dynamic resource request, and transmitting the results to the resource intermediating modules.