More and more ontologies are emerging across bioinformatics domains to represent and define domain knowledge, such as gene ontology, anatomy ontology and disease ontology. To integrate these heterogeneous ontologies is becoming critically important for applications utilizing multiple ontologies. Because the entities described in the ontology often overlap with other entities in other ontologies, a mapping between two corresponding terms is required and becoming the key to the integration of heterogeneous ontologies. Based on a previously developed schema matching algorithm, this paper presents a framework for mapping and integration of heterogeneous biomedical ontologies. The framework detects possible false mappings intelligently and provides intuitive interfaces for users to customize mappings and for applications to integrate and access biomedical ontologies.