Developing query rewriting techniques to resolve XML data integration Problems based on a static analysis.

融合是一种已知的技术，用于消除不必要的中间结果，所述中间结果被创建但随后在计算期间被消耗。融合对于许多查询语言（如SQL）的查询优化非常有用，但它仍然是XQuery优化的一个挑战。这是因为XQuery具有更复杂的语义;它是上下文敏感的，并且需要保持文档顺序。在这项研究中，我们提出了，据我们所知，第一个XQuery融合，可以同时处理文档顺序和上下文的XQuery表达式。更具体地说，我们精心设计了一个上下文表示的XQuery表达式的基础上的杜威顺序编码，开发了一个上下文保持有序树的XML存储的静态仿真的XQuery融合，并证明了我们的融合是正确的。我们的XQuery融合已经实现，并成功地应用到多步骤模式映射，其中融合是特别必要的，以减少冗余节点创建的执行成本。

Fusion is a known technique for eliminating unnecessary intermediate results that are created but then consumed during computation. Being useful for query optimization for many query languages such as SQL, fusion remains as a challenge for XQuery optimization. This is because XQuery has more complicated semantics ; it is context sensitive and requires preservation of document order. In this research, we proposed, as far as we are aware, the first XQuery fusion that can deal with both the document order and the context of XQuery expressions. More specifically, we carefully designed a context representation of XQuery expressions based on the Dewey order encoding, developed a context-preserving XQuery fusion for ordered trees by static emulation of the XML store, and proved that our fusion is correct. Our XQuery fusion has been implemented and successfully applied to the multi-step schema mapping, in which fusion is particularly necessary for reducing execution cost of redundant node creations.