Research And Implementation Of Distributed RDF Data Parallel Reasoning Method

RDF is a technical specification for marking the World Wide Web.It can express and describe the structure and content of network resources in a rich way.The RDF and OWL standards in the Semantic Web have been widely used in various fields.Due to the geometric growth of data in recent years,an efficient parallel RDF data inference scheme has been designed,and the important information hidden in the data has been mined.A problem that needs to be solved urgently.The main research content of this paper is parallelized inference algorithm of streaming RDF data.Combining Spark platform and Redis memory database,using MapReduce computing framework,we implement flow inference algorithm based on OWL Horst inference rules and extend the flow of OWL DL construction operator.Inference algorithms,and apply the algorithm to actual research projects.The main contents of the paper are as follows:Firstly,the PSRH algorithm is designed for OWL Horst rule.This algorithm combines the characteristics of Redis,and according to the characteristics of the rules of the pattern triples,the rule connection variable relationship table is designed so that the active rules can be quickly identified and It makes inferences;then the repeated triples are processed and saved in real time for the next iterative calculation.Finally,experiments prove that the algorithm is efficient and accurate in processing large-scale RDF data inference.Then,for the problem that the PSRH algorithm cannot handle the OWL DL construction operator,an inference rule and an inference algorithm PSRD for the extended OWL DL construction operator are designed.Through the transformation of matching rules,the reasoning of derived rules,and the construction of complex rule connection variable relation table,the inference rules based on OWL DL construction operators are realized.Finally,experiments are designed by designing a distributed computing framework.Experiments show that the PSRD algorithm can effectively extend the reasoning of the derivative rules of the DL operator.Finally,this dissertation applies the research of distributed RDF data parallel inference method to optimize a city's security risk management and intelligence platform.Firstly,the ontology construction method is used to construct the risk data ontology file,and the existing risk data is transformed into RDF data suitable for inference calculation according to the definition of the ontology file.Then the flow parallel inference algorithm proposed in Chapter 4 is used to achieve risk information inference.analysis.The RDF data parallel inference method for streaming data and the streaming parallel inference algorithm extended by OWL DL construction operator proposed in this paper are good for the inference of massive data for OWL Horst rules and the practical application reasoning analysis for OWL DL.Reference significance.