Study on a New Structural Join Algorithm for XML Query

With the problem that the cost is too much for temporary sorting or creating index aimed for there is no index or the index is out-of-order of input data sets in most XML structural join methods, the classical Stack-Tree-Desc algorithm and B+ tree index optimization algorithm are analyzed. A new strategy that isnpsilat limited to any external index is proposed and implemented. The experiment results indicate that it out performs Stack-Tree-Desc algorithm.