A horizontal fragmentation method based on data semantics

The need for distributed database systems increases day by day due to the need of running organizations from different locations. The efficiency and performance of distributed query processing depends on fragmentation and allocation method used. Usually, the fragmentation solutions are based on empirical data and analyzing query patterns. These methods can only be applied for the existing distributed databases and cannot be applied during initial database design. Also, most of the existing methods does not consider the semantics or relationships of the data being fragmented. Better fragments can be produced if data semantics are considered during fragmentation. Related data can be obtained from dataset using clustering technique. In this paper we propose a new horizontal fragmentation scheme based on clustering. Performance and time improvements are found while performing with clustered fragments.