Measuring and Comparing Semantic Structure of Ontology

Ontologies which are already constructed can be reused. Ontology reuse offers great benefits by measuring and comparing ontologies. A fundamental operation in ontology reuse is to compute similarity and dissimilarity among ontological entities such that one can establish certain level of correlation between ontological entities used in different ontologies by predefined measures and semantic comparison. Using Graph Derivation Based Approach, semantic structure of two or more ontologies can be compared and measured. In this paper, the novel method called as Graph Walk Based Method is used to compare and measure the semantic structure of the ontologies. This method compares the similarity with polysemy and toponymy. It re-rank the path covered in graph walk. It integrates the textual and structured elements of the data. The graph representation is modular and it is straightforward to extend the graph schema to include other information sources available, such as concepts that represent activities, relations that correspond to domain, etc. While the graph walk and weight tuning effectively model concept associations, re-ranking allows further flexibility in adapting the generated similarity measure to different tasks using high-level and task-specific information.