SMDPA: A SIMILARITY\|BASED MULTIDIMENSIONAL DATA\|SPACE\|PLACING ALGORITHM

This paper describes a new multidimensional data\|space\|placing algorithm, SMDPA. The algorithm can get the access relation between data hypercubes by using their priori frequency and similarities. Consequently, it can place the data hypercubes efficiently even if they have different accessed frequency. The simulation result shows that the SMDPA algorithm may obtain better performance than traditional algorithms do.