Research And Implementation Of Data Fusion Method Based On RDF

With the continuous development of Internet technology,more and more field data information posted on the Internet,the amount of data on the Internet is becoming more and more huge.In order to enable decision makers to find more accurate data information in large data,and make a correct judgment,data fusion technology becomes more and more important.In order to realize the data fusion,we need to unify the management of the data,W3 C proposes the RDF resource description framework which uses the three triples to express the data.In this paper,we use R2 RML mapping framework,SPARQL query language and Jena framework to realize the data fusion operation,which based on the three triples form of RDF.According to the characteristics of data fusion in the cloud computing project,a three layer data fusion model is designed to realize the data fusion in the talent database and the achievement database.In this paper,the data fusion method which based on RDF is studied,and some challenging problems encountered in the research process are analyzed and discussed.The main contents are as follows:1.Change the relational database to a RDF file.Introduces the basic storage structure and syntax representation of RDF.This paper analyzes the operation of converting data into RDF before data fusion which based on RDF,that is mapping from relational database to RDF data and mapping tools.2.SPARQL query language.In the RDF file,the data which needs to be fused needs to be matched,the RDF query matching technology based on the relation?based on the basic three triples and the RDF query matching technology based on the graph are compared and analyzed.Using the graph-based SPARQL query language to query the fusion data,and introduces the basic structure of the SPARQL query language and the basic syntax of the query language.3.A data fusion model is designed for the characteristics of data fusion in science and technology cloud project,designed a specific relational database to RDF file mapping rules.In the process of fusion may occur in the conflict of the value of the property,we need to deal with the conflict of the value of the property.Finally,the Jena framework based on the platform of Java,can read the RDF file,and create the model,through operate the model of the three triple,and then to achieve the operation of the RDF file.The RDF file is finally generated after the data is merged.Through the research results from this paper,we can find that RDF-based data fusion method can effectively combine the two triples in RDF files to form a triad representation through fusion operation.This can effectively reduce the number of recurring data,improve the decision-makers on the need data to make accurate determine,and provide more comprehensive and comprehensive information to the relevant users.In this paper,RDFbased data fusion method provides a new idea and solution for realizing data fusion.