Ontology mapping is the key technology to solve the heterogeneous problem among ontology, and concept similarity computation is a crucial step in ontology mapping process. In view of the problems existed in ontology mapping, an improved multi-strategy ontology mapping model is proposed. The most related concepts are filtered based on similarity of concept names. Then we calculate concept similarity by combining property-based similarity, structure-based similarity and instance-based similarity. Finally, we use the test set called benchmark provided by OAEI to test the performance of our mapping algorithm. The experimental results show that this method improves the recall and precision of ontology mapping while maintaining mapping efficiency and currency.