基于Hadoop的大规模RDF语义数据应用平台

Application Platform for Large-scale Semantic RDF Data based on Hadoop

随着语义网技术不断发展,传统推理引擎技术对大规模RDF(Resource Description Framework)数据的高效存储和语义查询存在着计算性能差和扩展能力不足等问题。针对这些问题,引入大数据处理技术Hadoop以及No SQL存储技术,构造一个大规模RDF语义数据应用平台,并且设计了基于SPARQL(Simple Protocol and RDF Query Language)查询技术的节点资源扩展算法。平台充分利用MapReduce技术提高了数据的存储和查询性能。实验以DBpedia的语义数据为案例,验证了平台的可行性和有效性。

With the development of semantic web, conventional rule-based engines meet the bottleneck in computing performance and scalability for storage and semantic query. To address these problems, application platform based on Hadoop and NoSql for large-scale semantic RDF （Resource Description Framework） data was presented. Algorithm based on SPARQL （Simple Pro- tocol and RDF Query Language） was also designed for extending resource node. Performance of storage and query was improved by taking advantage of MapReduce. The feasibility and effectiveness of the presented platform were proven through the case of DBpedia semantic data set.