<![CDATA[The required time to run a query on a database which possesses a large number of attributes and rows highly depends on the number of attributes accessed. An optimal query will only access necessary attributes. A vertical fragmentation of a table helps in isolating the frequently accessed attributes with those which are less frequently used.In this research, we develop an application to help the comparison of query execution time on a normal and vertically fragmented table. We use a Bond Energy algorithm to form the table’s vertical fragmentation. In this algorithm, the fragmentation process has to fulfill the completeness and disjointness conditions.We perform and test this application by using 40 queries from “Sistem Informasi Perencanaan Universitas Udayana”. Our test results show that a vertical fragmentation using Bond Energy algorithm increases the speed of query execution on the following: queries which access only several attributes from a table and queries with arithmetic operations such as sum or average. For insert and update queries, vertical fragmentation has no significant impact.]]>
Copyrights © 2012
