Efficient Semantic Information Retrieval System from Relational Database

Now a days, the large amount of data available on the internet is in the form of HTML documents. Most of the web applications (about 75%) using a Relational database is available as a source of information. In the Semantic Web the data can be represented in the form of XML, RDF or OWL. The Ontology is a structured data on the Semantic Web to organize and share the knowledge about the specific domain. In this paper, a framework is developed for generating Ontology and extraction of knowledge from the generated information source. This framework has two major phases, first is to generate the Ontology from the existing Relational databases using set of construction rules, and second is to extract the knowledge from the generated Ontologies. The mapping process is between the Relational schema and the Ontology schema and the Ontology is represented in Web Ontology Language (OWL). The Relational database has set of finite relations, attributes and values for attributes. The Ontology has a finite set of classes, properties and instances and also has some semantic properties like symmetric property, inverse property, object property and data type property between the classes and instances. The information is extracted from Ontology using SPARQL query language. This framework provides the efficient results for information retrieval from the Ontology.