摘要：Ontology mapping, based on similarity calculation,aims to find the similar concepts among different ontologies. Inorder to mathematically represent concepts, it is common torepresent all information of a concept in a fixed dimensionalvector. Therefore, the similarity calculation can be convertedinto a distance calculation between vectors, and the smallerthe distance is, the larger the similarity will be. In this paper,the low-rank matrix learning strategy is used to obtain thecorresponding ontology mapping strategy. The core idea of themethod is to control the upper bound of the distance of thesimilar vertex pairs in the sample and the lower bound of thedistance of dissimilar vertex pairs. At the same time, the rankof the matrix is integrated into the optimization conditions. Theeffectiveness of the proposed ontology trick is illustrated by theconstruction of ontology mapping on three ontology data.
