As a semantic analysis and calculation model,ontology has been applied to many subjects. Numerous ofmachine learning approaches have been employed to the ontology similarity calculation and ontology mapping. In theselearning settings, all information for a concept is formulated asa vector, and the dimension of such vector may be very large incertain special applications. To deal with these circumstances,dimensionality reduction tricks and sparse learning technologiesare introduced in ontology algorithms. In this paper, we raisea new ontology framework for ontology similarity measuringand ontology mapping. We construct the adjoint ontologygraph by means of index set of ontology vector. The optimalontology vector is obtained in terms of Lagrangian relaxationapproach. Finally, four experiments are presented from variousperspectives of different fields to verify the efficiency of thenew ontology framework for ontology similarity measuring andontology mapping applications.
展开▼
