基于本体的XML数据集成的查询处理

数据集成在很长一段时间一直为人们所关注．随着XML逐渐成为Web上发布和交换数据的标准，研究人员提出了基于本体的XML数据的语义集成方案．此文对于这样的方案进行了形式化，提出的算法NaiveMaxQEP解决了最大查询计划的生成问题；并基于非完整性角色个数有限的假设，对于查询执行计划生成算法进行了优化，提出了0ptMaxQEP算法；为支持高效的分布式资源整合，文章还给出了网络代价优化算法NetoptQEP，基于网络代价对输出的执行计划进行了优化．

Query Processing for Ontology-Based XML Data Integration

There has been a significant focus on data integration for a long time. Recently XML has become the de-facto standard for publishing and exchanging data on the Web. And researchers presented various integration schemes of XML data, but XML itself is not very appropriate for describing the global schema. One of the schemes which uses ontology to describe global schema gains our research interest. Because the original research work did not really present an algorithm to generate maximal query execution plan, after formalizing the scheme, a naive algorithm named NaiveMaxQEP to generate the maximal query execution plan is provided in this paper. And then an optimized algorithm OptMaxQEP is presented based on the assumption of limited number of incomplete roles in the global ontology. Finally, in order to support efficiency of distributed resource integration, an optimization algorithm NetOptQEP working on the generated maximal query execution plans to produce network-cost optimized plans is also presented.