Computing and Informatics

As XML has been established as the standard for data exchange not just on the Web but among heterogeneous devices, systems, and applications, effective processing of XML queries is one of core components of ubiquitous computing. Most of the mobile/hand-held devices deployed in ubiquitous computing environment are still limited in memory and processing power. An effective query processing is required when the source XML document is of large volume. The framework of fragmenting an XML document and streaming the XML fragments for query processing at the mobile devices has received much attention. However, the main focus was on the memory efficiency to cope with the memory constraint in the mobile devices. Query processing time might be compromised in those techniques. Since the processing power is also limited in the mobile devices, the time optimization deserves attention. We have found out that the query processing time is significantly affected by how the source XML document is fragmented. In this paper, we propose a method of XML document fragmentation whereby query processing gets efficient in time while the size constraint for each resulting fragment is satisfied. Through implementation and a set of detailed experiments, we show that our proposed method considerably outperforms other methods.