Efficient Multi-Query Evaluation over Compressed XML Data in a Distributed Environment

Aoying Zhou,Juzhen He,Wilfred Ng
IF: 2.656
2007-01-01
Journal of Database Management
Abstract:With increasing dissemination of XML, multi-query processing has become a practical and meaningful issue to resolve. However, the verbosity of XML data causes inefficient query processing and high network bandwidth consumption. This paper addresses the problem of evaluating a heavy load of subscribed queries as a whole (or simply multiqueries) over compressed XML data in a distributed environment. We propose a holistic approach that evaluates complex queries over a compressed document and directly forwards the compressed results to clients. We first introduce a new rewriting technique, which is used to decompose and reorganize a complex query into its corresponding Structure of complex XPath (SXP). Then, multi-query evaluation can be performed based on the containment relationships between the queries. The containment relationships are exploited by a global data structure, Structural-Query-Index Tree (SQIT). SQIT is an efficient structure that supports prefix sharing among the submitted queries, reserves all result locations as indexes after evaluation, and establishes a sequence of result publication. The analytical and experimental results demonstrate that the proposed approach can obtain higher query processing efficiency than traditional ones, and the bandwidth cost can be saved substantially. Thus, our approach is particularly suitable for large numbers of geographically distributed users who need to access a massive amount of correlated XML information in a cooperative distributed environment.
What problem does this paper attempt to address?