Abstract: A method is provided for generating a destination address in response to a plurality of oral inputs by a user. The user is prompted to orally input a primary geographical descriptor of the destination address to the in-vehicle navigation module. Primary geographical utterance data is generated in response to an oral input by the user. The primary matching voice templates stored in the in-vehicle navigation module are compared with the primary geographical utterance data. The secondary voice templates associated with the secondary geographical descriptors within the primary geographical descriptors are retrieved from the remote navigation server. The secondary geographical utterance data is generated in response an oral input by the user. The stored secondary voice templates are compared with the secondary geographical utterance data. The destination address is generated in response to matching voice templates.