| XML is the preeminent data exchange format on the Internet. Processing XML data has recently become a widely recognized area of interest both in research and in industry. To organize and manage XML data efficiently, different storage approaches have been proposed. Among them, relational databases and file systems are the most popular ones. With the development of Web services, XML is also increasingly served by remote Web service providers. Therefore, there exist different kinds of XML data sources.Despite current research achievements on XML query processing and transformation, there are still many open issues to be explored. This work investigates some issues related to query processing, data transformation and integration of different kinds of XML data sources, which include: XML query processing in relational databases, especially, building path indexes to speed up query evaluation; XML document querying and transformation while guaranteeing the correctness of their result types; XML data transformation without extra memory buffers required; integrating XML data returned by different Web service providers, etc. The main contribution of this work can be summarized as following:· In order to speed up query evaluation for XML data in relational databases, build part of path indexes which can bring much benefit to query evaluation. Making use of the statistics of XML data and the characteristics of the target application, a cost model is presented for estimating the space consumption of indexes and their benefit to XML queries. Greedy algorithms are adopted to select some map indexes to be built. The experimental study demonstrates that query performance gets significant improvement over the case without indexes while only consuming disk space of modest size.· For querying XML text files, attribute grammar is extended with XML queries to obtain a new XML query language (XTG). To some degree, XTG guarantees that the result document will be correct. That is, the result...
|
PDF Full Text Request