Research On Parallel Query Of XML Stream Data Based On Pushdown Tansducers

The social networking,network monitoring and financial applications have a need to query high rate streaming of XML data?The existing XML stream data query methods do not give full play to the computational advantage of multi-core processors?Data parallel query methods improve processing efficiency by using multi-core CPU? Data parallel query includes two processes:data processing and link processing ? In the process of data processing,multi-block data is processed simultaneously,and the link processing connects the processing results of the data processing process,reconnect the truncated block?This approach makes full use of multi-core computing resources,so that the utilization ratio of the processor is greatly improved?Automata as an effective means to deal with semi-structured data,are often used to process XML data?Push down conversion machine as a special type of automaton,it not only can be used to handle XML data,and its internal stack structure can preserve the important information in the processing,so it can combine with the data parallel querying to improve the processing efficiency further?When processing XPath queries with predicates,the usual approach is to rewrite XPath,resulting of getting a set of sub queries?These sub queries are not completely independent of each other,often each other depends on?As a result,the link operation increases the link matching processing between sub queries(the processing time is often longer than the processing time of the sub-query),resulting in low efficiency?In order to reduce the processing burden in link parse,so as to enhance the efficiency of the link query,XPath query mode is converted into multiple ordered tree pattern queries(each query is effective and independent,the matching results of these query equals to XPath query mode)?The link processing stage only needs to carry on the link processing of a single query result,eliminates the connection processing between the queries,and reduces the processing time of the link stage?Through the effective combination of data parallel querying and pushdown transducer,the multi-core computing resources are fully utilized?At the same time,the ordered query tree pattern as the matching mode can reduces the data link processing expense,so the processing efficiency significantly improved?...