针对可扩展标记语言(XML)查询中具有嵌套OR谓词的复杂小枝模式查询处理,提出一种基于路径连接的查询方法.该方法以路径为分解粒度,结合分支扩展(AOBE)的概念,通过路径连接过程实现对复杂小枝模式查询的整体处理.为了进一步提高算法效率,在已有研究的基础上挖掘相应的优化规则,利用索引跳过那些明显不参与连接的元素的访问和计算.与已有算法相比,基于路径连接的查询方法大大简化了复杂小枝模式查询处理过程,只访问查询叶节点对应的元素,可以显著减少结构连接的操作数目和扫描元素的个数.实验结果表明,该方法能够有效地改善复杂小枝模式查询处理的性能.
A novel path-joins based method was proposed to support efficient processing of complex twig pattern queries with OR-predicates of extensible markup language(XML) queries.The method processed the complex twig pattern matching in a holistic way based on the concept AND/OR branch extension(AOBE) and path-joins by dividing the twig pattern into individual paths.Then an index-based algorithm was proposed to efficiently skip useless elements and avoid unnecessary computations.The path-joins based method simplified the complex twig pattern queries processing compared with the existing algorithms.The method only accessed the labels of leaf query nodes,thus the I/O and CPU costs were greatly reduced.Experimental results demonstrate that the method is more efficient than previous approaches.