Comparison on Performance of Text-Based and Model-Based Architecture in Open Source Native XML Database

A large number of organisations today choose XML as standard format for business transactional documents. Such documents can be natively processed by open source Native XML database based on either text-based or model-based storage architectures. This paper presents a comparative analysis of these two architectures in terms of storage and query performance in handling XML documents. In specific, the paper compares Berkeley DB XML, eXist and dbXML performance. Three dependable documents were deployed for testing purposes and investigation was directed to focus on the speed of executing storage instructions and speed of executing querying instructions. In data analysis, line chart and histogram were utilised to reflect the trend of the processing performance of the database graphically. The result reveals that text-based architecture outperformed in storage operation whereas model-based architecture performed better on query operation.