Research Coding And Query Algorithm Of Graph Structure With Cycle

With the developing of xml technology. The area of management and query technology on cycle graph-structured attracts more and more people to research. And the technology to deal with the question of the cycling on graph-structured is becoming popular, how to provide a code algorithm to process circle base on graph-structured and design a query algorithm has became main direction on XML.Because of the interval encoding is not apply to the research the graph-structure, and the interval encoding must be to judge mass data from the graph usually produce large middle result, therefore seriously reduce query efficiency.It not to be used to solve the relationship of ancester or parents between nodes on directed cycle graph, so we need to find a new algorithm, this algorithm is solved by the relationship database, as to improve coding rate and saving space. Decrease the encoding time. For the query algorithm on cycle graph, most of the algorithm is development base on the tree-structure,so it is not apply to query on the graph-structure.In this paper, a new encoding algorithm on graph-structure are presented, and this method use the depth finding strategy and set up the flag to deal with the question of display circle on graph-structure. Avoid storing a great deal of middle result, and it can determine the reachable relation between any two nodes in very short time. Based on this conclusion, this paper provide another query method, which flowing this method traversing all nodes of query graph once in graph, and judge reachable relation, at last get the result of the query. Query result is represented by code of nodes. with small storage space and have a high efficiency.In this paper, research the query technology base on the CUDA algorithm and give up the interval coding and subgraph query, use the triple to solve the question of accessible query, put the result in the list table, compare with the HGjoin query that store the result in Alist and Dlist tables only in half the space the same data, because abandoned the interval coding do not need to repeating the result sets, to raise in rate.