On Ontology Learning Algorithm of Pairwise Ranking
Zhu Lin-l · Journal of Southwest China Normal University · 2013
As a structured data model,ontology has been widely used in knowledge representation and information processing,and it becomes a hot topic of computer science in recent years.In this paper,new ontology similarity calculation and ontology mapping algorithm based on the pairwise ranking learning method have been discussed before a model is given by means of Mahalanobis distance function.The optimal solution has been obtained in views of gradient descent strategy.Thus,each pair of vertices in ontology graph or mutil-ontology graph is mapped into real number,which represents their similarity.Two experiments show that the new algorithm has a higher efficiency for specific applications.