中文摘要：

针对现有方法计算SLCA语义时存在冗余计算问题，提出了一种基于列存储的倒排索引，并结合哈希查找，以自项向下的方式查询处理的算法TDCOL-HS，来避免现有算法“公共祖先重复处理”的问题。算法以最短倒排表作为处理对象，将检测给定结点是否包含其他关键字的操作转化为哈希查找操作，其时间复杂度为O（mxL），最后通过比较各种指标，从不同角度对算法的性能进行了验证．

英文摘要：

Considering that existing methods suffer from redundant computation when processing XML keyword queries for SLCA semantics, we propose an efficient algorithm, namely TDCOL-HS, which processes the given query based on column storage and hash probe operation to avoid the problem of repeatedly computing common ancestor nodes. It takes the shortest list as the working list, and transforms the operation of testing whether a given node contains other keywords into hash probe operations, therefore, achieves the time complexity of O（mxL）. The experimental results demonstrate that the performance benefits of our methods in adding key word search on XML data.