Performance of relational databases versus native XML databases
The popularity of Extensible Markup Language (x.mii) has grown with the rise in use of the World Wide Web and it has become a standard means of data exchange. This development has created the need to store and manage XML documents efficiently and hence the creation of the Native XML Database (NXD). Claims have been made to support the efficiency of this technology with the storage and management of document-centric data over the prevailing Relational Database Management System(RDBMS). This research thus aims to discover evidence supporting the conclusion that there will be a significant difference in performance between a RDBMS and a NXD with the storage and management of document-centric data. Performance is based on the required execution time for a set of queries that implement the possible types of disk access. The query set is run on an instance of each database type and query-execution-time is collected and analysed to provide a measure of performance. This paper concludes there is a significant difference in performance between the database types and further concludes that performance, if based on query-execution time, is dependent on specific task.
Degree Discipline: Information Science
Keywords: XML databases; relational databases; extensible markup language; manage XML documents; database efficiency; RDBMS; NXD; document-centric data; query-execution time,
Research Type: Dissertation