Research Of Distributed Storage Of Massive RDF Data

With the rapid increase of Web data and various network resources as well as the rising and development of the Semantic Web, Massive RDF (Resource Description Framework) Data storage becomes a hotspot in the current filed of Web Data Storage. In this paper, the author proposes a scheme which can store Massive RDF Data in a distributed solution, on the basis of the further study and research of several popular distributed storage framework. The scheme can also store data efficiently and collaborative in linux cluster based on Hadoop Database (HBase).Firstly, the author analyzes and compares some kinds of distributed storage framework in a deep way. Combining with the characteristics of this project, we select a open source framework Hbase (Hadoop DataBase) on which we can store both sparse and Massive RDF data in Hbase table. Not only dose it solve the query problem of low efficiency combined in traditional relational database, but also can speed up processing speed with the MapReduce algorithms on the distributed platform.Secondly, the paper analyzed the open-source framework Hadoop and the Semantic Web framework Jena, and then parses the RDF/XML documents into RDF models, creates the RDF models and made a semantic analysis of some models. The author rewrites almost all of the storage underlying code in Jean Framework so that it can migrate Massive RDF Data to the distributed platform. The author also rewrites some parts of the Jena model parsing code and query code in order that it can give full play of distributed processing. Meanwhile, Accelerating query efficiency and processing speed will laid the groundwork of Semantic Web.Finally, the author overviews the research results of the whole project and proposes the object and foreground of the research combined with the own research interests.