Mining large data sets on grids: Issues and prospects