Research And Implementation Of RDF Keyword Expansion Query Algorithm Based On Bipartite Graph

As more and more RDF data sets in many fields have been published, the RDF data retrieval and reuse is becoming more and more important. Existing RDF data query, however, where research there are still some shortcomings: although the RDF data query method based on graph model can maintain the semantic information of data, but it brings the query response bottleneck problem with the increase of RDF graph scale; Because of keyword query without considering the problem of word ambiguity, the keywords cannot be matched, and the query results cannot satisfy the user’s query intention; In the face of massive RDF data query processing, the traditional query methods are all in performance bottlenecks. In view of the above problem, through to analysis and study of a large number of literature materials, this paper put forward to adopt RDF bipartite graph and extend query keywords, thus efficiently implements RDF data query. The article mainly works as follows:(1) The bipartite graph model of RDF data is put forward. On the basis of the traditional RDF data graph model, the model uses vertices said RDF triples of the subject, predicate and object, increases the query to the predicate, and completely retains the semantic information of RDF data; A theorem is proved that the anti-symmetric adjacency matrix and the exponential matrix of RDF bipartite graph can construct the query result sub-graphs that contain query keywords; By introducing some relevant theoretical knowledge of the cut set, the theorem is proved that the same vertex sets between the sub-graphs after segmentation are cut sets, to ensure the effectiveness of the partition algorithm theoretically.(2) The article puts forward RDF keyword expansion query algorithm based on bipartite graph. The algorithm gives the synonym expansion of query keywords by using Word Net dictionary, and through the use of the anti-symmetric adjacency matrix characteristics of the bipartite graph, the process of query processing is optimized. Experiments show that compared with traditional methods, the proposed algorithm can eliminate the words ambiguity, improve query accuracy and reduce the query response time.(3) The article puts forward the RDF data segmentation algorithm based on bipartite graph. In order to achieve parallel query algorithm, improve the query efficiency of huge amounts of RDF data, through the analysis of the key factors for graph cuts, to make RDF data segmentation principles based on bipartite graph, to theoretically prove that the same vertex sets between the sub-graphs after segmentation are cut sets, and to give the RDF keyword expansion query algorithm based on bipartite graph of parallel implementation combined with segmentation algorithm.The experiments verify the effectiveness of RDF query keywords extension algorithm and the segmentation algorithm based on bipartite graph proposed in this paper, and give the performance analysis of the query algorithm before and after parallelization.