Introduction:Ontology is basically a formalspecification of conceptualization and that is shared. In semantic web ontologyconsider as a major pillar the major reason for this is a sense of knowledge ofa domain. Ontology mapping can be define as the integration andmerging of ontology. Integration of ontology means that design a new ontology whichuse again other available ontologies and merging of ontology meanscollaboration of different ontologies into a single one make union of them.
Mapping of ontology is based on some factors which play a vital role. First ofall considering number of different properties of ontologies like as syntax,semantics and structure. Even many researchers’ works in this area, ontologymapping techniques are not still perfect and there is possibility for the improvementin this. While the major purpose of the ontology mapping is allows different schemas or ontologies todetermine information that is same although it is based on different schematicsyntax or a structure. In this just discuss few of them.
OntologyMapping Techniques:1. ARTEMIS (Analysis andReconciliation Tool Environment for Multiple Information Sources).It is used tomanage the access as well as the availability of dynamic collections of thedata sources along this it also provide the supports of the creation of queriesstep-by-step for user. The major aim of this is solving the problem of independent data stores.According to (Castanoet al., 2001) it involves three key steps:1.
Analysis ofschema elements2. Clustering ofschema elements3. Constructionof global viewsEuropean Commission FP6 funded ARTEMIS Project and developed a frameworkAMEF stands for Artemis Message Exchange Framework. It is used to provide themeaningful information.
The initial task of this is mapping of source ontologyinto target message ontology through a mapping tool which produces a mappingdefinition and then this used to convert the source ontology message intotarget message automatically. Inorder to represent the matching between source and target classes based ontologies,here defined four mapping mechanism which are:1.EquivalentTo2.SimilarTo3.IntersectionOf4.UnionOf Twoclasses that are indistinguishable and mapped through EquivalentTo mechanismand SimilarTo mechanism involves that classes which have overlapping thecontent.
Inproperty mapping mechanism, similar classes are further related through theirdata type properties and object. HL7 version 2 ontology is defined to be similar to HL7 version3ontology. Using “ObjectPropertyTransform” mechanism to define the mapping of “hasValue”and “hasQuantity” object properties. While “IntersectionOf” creates thecorresponding instances of the target class as the intersection of the declaredclass instances and “UnionOf” is similar. Using KIF “KnowledgeInterchangeFormat” to define the instances of thesource ontology that makes instances of the target ontology.
OWLmt shows thepath of classes that are linked by the object properties in such a way that eachand every time a path defined in the source ontology means inputPath isencountered in the source ontology instance, the path which is defined fortarget ontology means outputPath is generated in the target ontology instance. Paths can be define as triples in KIF format and executed through theOWL-QL engine. Assuming the path that is define in the source ontology In the target ontology following path correspondenceOWLmt constructs the specified paths among the instances of the targetontology in the execution step based on the paths defined among the instancesof the source ontology. Mapping Definition 2.
COMA COMA stands for combined matching. Authors presentthe COMA mapping technique in (Do and Rahm, 2001), this is a system for combiningdifferent ontology mapping methods which is based on the different situationand source schemas. In 2005 COMA++ was introduced in which re-implementation ofthe original COMA design due to this improve performance, functionality and alongwith this it also provides a GUI that simplifies the definition of match strategies.
In COMA Match Process