Navigation system with speech recognition

摘要

Method for production of a data structure, especially for phonetic transcriptions relating to place names for a speech operated navigation system in which from a whole quantity (G) of data (4) subsets (1a-d, 2a, 2b) are formed, the elements of which fulfill at least one criterion. The subsets are combined to produce an intersection set the element number of which does not exceed a predefined value. An Independent claim is made for a speech recognition device for recognition of spoken geographical names.