Ontology mapping algorithm based on dimensionality reduction theory
Wei Dong Gao · Journal of Northwest Normal University · 2011
This paper aims to establish a mapping by multiple ontologies to achieve interoperability between multiple ontologies.Each vertex in multiple ontology graphs can use a vector corresponding to the vertex's information.The connected ontology graph is obtained by calculating the Euclidean distance for each pair of vertices.Using spectral clustering dimension reduction methods,the dimension for each vertex is reduced to one-dimensional data.The ontology mapping is given by measuring the difference of values which reflect their similarities.Experimental results show that the new algorithm can solve the problem of the literature and is effective for mapping in ontologies of large differences.