Research On Xpath Query Processing Over XML Data Stream

With the development of computer networks and communication technology, the researches on data stream have attracted considerable attention. Practical applications of data stream manage system(DSMS) can be found in many areas such as financial analysis, sensor networks, traffic information systems, mobile object tracking, network data monitoring, etc. On the other hand, XML data formats has become the standard of network data exchanging due to its inherent advantages and practical application requirements for efficient network communication. The data stream has been expressed in the form of XML formats. For example, in the environment of Web Services, a continuous XML data stream is produced when mass data be exchanging between loosely coupled systems. Thus, the theories and technologies related to XML data stream has become a central issue in the data stream research field.XML data stream is generated and transported in high-speed, and has such features as ordering, real-time, uncertainty, one-pass scan, etc. Thus, how to efficiently processing the data stream, such as filtering, analysis and query, is the key issues of current research. At present, algorithms that has been proposed include automata based, index based etc. DTD can be of great help in improving XML query efficiency, however, most current algorithms do not use this effective resource.This paper focuses on the key techniques of query over XML data stream. Around the data stream query processing, the research content can be divided into two parts:XML documents stream filtering, XPath query processing on XML data streams. The techniques and theory involved includes DTD, functional dependency, automata techniques, indexing techniques, etc.The main work of this paper is:1. Proposed the documents filtering model and filter matching algorithm based on XML data stream. We discussed the applications of automata theory in the XML data processing. By extending the definition of DTD and tree automata, integrity constraint and functional dependency are introduced into query processing. The documents filter matching can be completed in one-pass scanning.2. Proposed a query processing model and the query algorithm based on XML data stream. A novel indexing method DBXSI is designed for XML data stream, structural information of DTD is introduced into the DBXSI index. And the stream query algorithm based on DTD and DBSXI is also proposed. By the matching between XPath tree automata and DTD tree, we can pre-mark the DTD nodes that not match the structure of XPath query. When the XPath query is running over XML data stream, the nodes and sub-trees which do not match the query can be skipped by the information of stream index to improve the query efficiency.