With the development of the Semantic Web, a global interconnected media is being generated, more and more RDF datasets are published on the Web. A DBMS that can store and query large RDF data is needed. To tackle the problem, the authors have designed and implemented a distributed RDF semantic data storage scheme. The scheme manages and queries RDF data effectively by loading and building triple indexing for RDF triples. The job includes designing and implementing the underlying RDF storage scheme, data pre-processing and loading. Also, the authors have devised a series of tests to evaluate and compare performances between a single-node Cassandra Database and a multi-nodes Cassandra Cluster with a 13m RDF Datasets from Dbpedia. Test results show that their scheme demonstrates its highly efficient storage, query and concurrent access performance.%随着语义Web的发展,越来越多的RDF数据发布到Web上,需要一个可以提供存储和查询功能的数据管理系统来对海量的RDF数据进行管理.针对上述问题,设计并实现了一种大规模RDF语义数据的分布式存储方案.该方案通过RDF数据装载和预处理,可以有效地管理海量的RDF数据,并通过构建索引可以有效地对大规模RDF数据进行查询.工作包括底层的RDF存储方案的设计与实现,数据的预处理与装载.同时,设计了一系列实验来评估和对比不同节点数目的Cassandra集群之间的性能,数据采用的是从DBpedia获得的13 million行RDF的数据集.实验结果显示,方案对大规模RDF语义数据的存储和查询具有性能优势.
展开▼
