Research On Seconde Index Technology For XML Data

XML as the new generation's data exchange standard, are getting more and more use in the network application. There are the massive XML data. How to carry on the effecitive inquiry to the XML data to process the hot spot which into the present studies. However, XML data is a semi-stucturized data. It has the characteristics of definition and expand. These characteristics have caused certain difficulty to the XML data query work.The index achievement accelerates to inquire the processing an important means that is playing the crucial role in many domains. Therefore, enhance the data using the index is an feasible means. This article in carries on the generalized analysis to the domestic and foreign research present situation in the foundation, has further conducted the thorough research to the XML index technology.Paper the index method which proposed to the recent years has conducted the study research, has analyzed the existing index characteristic and the good and bad points. Based on this proposed one kind uses the different index structure according to the way to carry on the inquiry the method. This method has division of simple path and the branch path, uses the compression tree index and the hash coding index separately carries on processing.The compression tree index used to processing the simple path to inquires in the process. This method not only has provided the succinct way outline in the group level, and has supplied the fathers and sons in the element level the relational detailed link informantion. When the element level mapping may a fast vist element father node, the group level may reduce the research space effectively. This index use replaced based on group's element quotation has carried on the code to the element, reduced the space which the index needed. The introduction improvement's platoon index and the reverse index speed up to the element judgment, more highly effective obtains in the index group the information. Takes the reference based on the group level's element, not only may cause the compression tree to become according to the group cluster arranges in inverted order the table, thus for arranges in inverted order the table and the group scope in group level. Thestructure index provides a better connection, but may also easy to carry the classification according to the group level. The close element will put in the same place in advantageous for the index.The main idea of the hash coding index is make each node correspondence a index tree node. In the index tree non-leaf nodes is depositing this node in the XML documents position.What the index tree saves is in the XML documents the father node to its subnode's path hash coding. In the index tree, leaf node is storing the XML data. The hash coding path index has the quite good extension, this index replaces in the beforehand index through the memory node's hash coding to save the node.Thus, it can reduced the storage space which the index file needs. Through the node carries on the code to the documents tree, trades the string of character attire for the integer, raises the inquiry speed. In the index introduction tradition database the level index's concept, has realized to the frequent use path direct access.This article proposed a level index. It can use the different indexed mode according to the path expression's special details. It is good at improve the efficiency of retrieval.