In order to deal with the need of sharing learning objects within and across learning object repositories most of the recent work argue for the use of ontologies as a means for providing a shared understanding of common domains. But with the proliferation of many different ontologies even for the same domain, it become necessary to provide mapping process to perform interoperability. The main key issue must be addressed is to define educational systems interoperability mechanisms to create a virtual learning space. Although many efforts in ontology mapping have already been carried out, few of them use resources properties to generate semantic relations between local concepts. Our approach uses inference rules to combine several matchers in order to discover mapping dynamically and to improve the results qualities.