Research On XML Index Technology

With the rapid development of Internet, XML has gradually become a new standard for information publication, data exchange on the Internet-known for its strong data representation capability , openness, extensibility and numerous other benefits. so many kinds of XML technology become a hear research focus. But with the extensive application of XML, the XML data to the dramatic increase in the XML file maintenance is a big challenge. To better XML document for easy maintenance and more on XML document for operation. People from the storage of the index, and then to inquiries made a positive and thorough study. These studies for the development of XML address the reality of many problems. Many outstanding research results have a certain scale application such as lorel. However, due to the diversity of XML data and the growing user demand inquiry, It is difficult to find a simultaneously applied to different data sources(such as Native XML Database and related Database) and handle inquiries common index structure. People use different indexing structure to meet under different circumstances demand. Present, research on XML index mainly concentrated in the node type and structure of records Abstract category two indexing method, but they have efficiency of index upgrading low and problems of dealing with Twig structure which will produce a large number of useless middle results . The above-mentioned factors reducing the performance of XML inquiring .In this dissertation, the above several issues are studied. The main novelty and contributions can be summarized as follows:1. MTB-Index supports dynamic updating of the index2. MTB-Index It uses linear model and path matching the inquiries into the match interval3. MTB-Index can solve efficiently the query with relation'//'4. MIF-Index occupied a small space, it can improve on the performance of memory 5. MIF-Index structure to deal with Twig no link, it avoids the shortcomings of dealing Twig Structure which must join...