Research Of Distributed Keyword Semantic Search Method Over RDF Data

The Semantic Web gives explicit semantics for data so that machines can better display,understand and process them.RDF(Resource Description Framework)is a framework for describing the resources on the World Wide Web,which is proposed by the World Wide Web Consortium(W3C).With the full development of the Linking Open Data project and the DBpedia project,many fields use RDF as the data model to publish their own semantic knowledge base.The structured query language is a standard query language for accessing semantic data,but the complicated query rules can't meet the query requirements of ordinary users.In order to meet ordinary users'needs for accessing the growing RDF semantic data,it is of great significance to put forward an efficient distributed search scheme for popularizing keyword semantic search and sharing semantic knowledge.Firstly,a distributed semantic search on large scale RDF data DKSSO algorithm(Distributed Keyword Semantic Search on Ontology)is proposed.It stores RDF ontology and RDF instance on the distributed database Hbase,and classifies large scale RDF instance data according to RDF data type to help locate and effectively narrow the search range.During the search,we first use RDF ontology to map keywords from the content level to the ontology semantic level,and build ontology subgraph for query keyword.And then the semantic score function is defined to score and rank ontology subgraph.Then,it uses MapReduce make parallel computation and preferentially search result subgraph for ontology subgraph with higher score on large scale RDF data graph until it finds Top-k results.Secondly,with the increasing application demand for online analysis over continuous data stream,data processing has changed from the static data to the dynamic data.In order to meet users' need for real-time search over RDF data stream,we put forward a distributed real-time search algorithm over large scale RDF data streams DKSSRS algorithm(Distributed Keyword Semantic Search over RDF Data Stream on Storm).It processes continuous RDF data stream in real time based on distributed stream computing framework Storm,and designs a distributed storage schema to store stream data and uses timestamp to distinguish the historical data and the new data steam.Then,it constructs real-time query storm topology task to receive and process real-time query request.Simultaneously,we establish query cache to store historical query result so that they can be reused to support incremental update query and complete the efficient real-time query.Finally,the experiment is performed on the benchmark dataset and the real dataset,which shows that DKSSO algorithm and DKSSRS algorithm both have obvious advantages in search efficiency and search effect compared with the existing search scheme.In addition,the experiment verifies that DKSSO algorithm and DKSSRS algorithm are both better scalable,and DKSSRS algorithm can process real-time query correctly and efficiently in a distributed environment.