In order to solve accuracy of Topic Maps of navigation and positioning for service clustering in the distributed software system, we proposed a navigation and positioning model of Topic Maps based on ontology relation (NPTMO)，and according to NPTMO, we build a service clustering model and semantics ant colony algorithm. Therefore, which implement navigation and positioning of resources from relation of Topic Maps. Firstly, we define ontology relation and calculate approach, and analyze associate characteristic of Topic Maps based on ontology; Secondly, we analyze and study to match and merge approach of Topic Maps, and employ tag relation of RDF, OWL-S, WSDL and XTM what to build navigation and positioning system,and build a service clustering model, on the basic of these context above description. We design a semantics ant colony algorithm (SACA) to calculate optimum solution of the model, and compared SACA with basic ant colony algorithm (ACA) showed that results of SACA was better than ACA. Finally, we define recall rate of ontology, accuracy of Topic Maps and efficiency of SACA to evaluate the approach showed that increased accuracy of Topic Maps connect resources.