The analysis for ripple-effect of ontology evolution based on graph

Abstract

Ontology is an important foundation for the Semantic Web and ontology-driven applications. While real network and application requirements are constantly changing, in order to adapt to these changes, the ontologies need to be updated in time. There are various types of semantic relationships, between the elements in ontology that the strength varies as well as their impacts of the process of evolution. To reflect them to the analysis of the ripple-effect of the ontology evolution, the common types of semantic relationships in ontology was extracted, and on this basis, proposed SRG (Semantic Relationship Graph) graph model in which the property-related semantic type substitutes the property and established the matrix and reachability matrix of the relationships among the elements in the ontology corresponding to this graph model. Then, the ripple-effect during the process of ontology evolution was analyzed via using matrix operations to calculate the comprehensive influence of the node in ontology. Experimental results show that the proposed method can provide accurate quantitative analysis for ripple-effect of ontology evolution under the premise of efficiency is possible.