Research On Distributed XML Query Processing

XML is a description language for semi-structured data. As a W3 C standard, XML is simple, cross-platform, self-descriptiveand widely used. The rapid growth of Internet data results in the continuous expansion of data size,data is distributed and stored in network based on actual demand. Distributed data storage and management are effective technological method for handling large-scale distributed data. As a de facto standard for Internet data representation, storage and exchange, XML is used in many distributed applications. XML data is naturally distributed and stored in network in applications such as information sharing, web service and e-commerce. Distributed XML query largely determines the performance of application system based on distributed XML data.Selection, projection and join are the most commonly used operations in distributed query. Selection and projection only involve local data, while join is an important operation for multi-data source query. Since different data sources are involved, data and message passing between sites may be required when join operation is implemented, thus increasing the communication cost and the total cost. Therefore, the communication cost caused by join operation has a great impact on the performance of distributed queryA query processing algorithm for tree-structured XML data is proposed based on the Semi-joins technology..In distributed XML query processing, the communication costs caused by the join operations in query involving different sits are critical for the whole query efficiency. The algorithm use the result of XML tree pattern query(Twig witness tree) as the base data, it could handle distributed queries containing join operations. Bloom Filter technology is used in the algorithm to reduce the amount of data transmission. The experiments show that the proposed algorithm could improve the performance and reduce the communication cost in distributed XML query processing.