The Research And Implementation Of The RDF Data Storage And Query Based On Graph

With the rapid development of computer and network technology, information systems are more and more complex. Distribution, heterogeneity, and large-scale of the information system become a challeng in data integration. Web data management and process are facing semi-structured data, large amount of data, slow querying speed, low efficient query, extensibility and universality. In order to satisfy the non-relational and heterogeneous data storage and processing requirements, this paper studies RDF data management and processing, put forwards the RDF data storage and query based on graph for improving the RDF data storage and query efficiency. First, this paper studies the RDF storage method based on hypergraph split. According to the RDF data characteristics, this paper put forwards weights generation strategy based on hypergraph hyperpoints and hyperedge and split RDF data graph method according to the weight and the smallest cutting set. Comparing with classic Metis and hash algorithms, paper analysis this method advantages and disadvantages. Secondly, we study the SPARQL oriented optimization method for generating a query plan, according to the statistical cost model, SP indexing mechanism and data storage method. To minimize middle of the query results each time according the query optimization sorting, we improve the query speed. Finially, we implement distributed storage, query and processing system of the RDF data based on Hadoop/HBase according upon proposed method. We compare to open source RDF system Sesame with experiment, the result show that our method is more effective to improve the RDF data query efficiency, and it is a good support system and method for RDF data processing.