| Summary: | 碩士 === 國立嘉義大學 === 資訊工程學系研究所 === 96 === Because XML documents have a tag structure that is easy to be interpreted by both human readers and computer programs, XML data format has been adopted for many cross platform applications. Most XML-related research in the past focused on improving the efficiency of various query processing methods, in which TwigStack [4] was the first holistic twig join algorithm. TwigStack uses stacks to reduce the size of intermediate results and to save the time of processing binary structural joins [1]. This research focuses on the subject of cached XML twig patterns, which are generated by an improved stack-based XML query processing method [6]. When processing an XML query, each query node is mapped to a stack and XML document elements are put into their corresponding stacks. If the data in the source database is updated, the cached twig patterns can be maintained using the updated path information. In many cases, it is not necessary to return to the source database for retrieving the updated information. Thus, incremental maintenance of the cached twig patterns can effectively reduce the server computation costs and the network transmission costs. Experiment results indicate that the proposed method is feasible and can save time.
|
|---|
