![]() ![]() ![]() |
![]() |
|
|
![]() ![]() ![]() ![]() ![]() |
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Return to Poster Papers Data fragmentation offers various attractive alternatives to organizing and managing data, and presents interesting characteristics that may be exploited for efficient processing. XML, being inherently hierarchical and semi-structured, is an ideal candidate to reap the benefits offered by data fragmentation. However, fragmenting XML data and handling queries on fragmented XML are fraught with challenges: seamless XML fragmentation and processing models are required for deft handling of query execution on inter-connected and inter-related XML fragments, without the need of reconstructing the entire document in memory. Recent research has studied some of the challenges and has provided some insight on the data representation, and on the rather intuitive approaches for processing fragmented XML. In this paper, we provide a novel pipelined framework, called XFrag, for processing XQueries on XML fragments to achieve processing and memory efficiency. Moreover, we show that this model is suitable for low-bandwidth mobile environments by accounting for their intrinsic idiosyncrasies, without sacrificing accuracy and efficiency. We provide experimental results showing the memory savings achieved by our framework using the XMark benchmark. ![]() ©2006 Association for Computing Machinery |