This paper discusses the problem of non-native pronunciation variants of foreign city names from the perspective of speech technology applications such as travel information or car navigation systems. In this application scenario, a broad spectrum of potential mispronunciations and pronunciation variants must be anticipated, ranging from minor phonetic shifts to heavily accented forms that hardly resemble the canonical forms provided in the phonetic dictionaries used for automatic speech recognition and synthesis. After highlighting some general properties of city names (as opposed to standard vocabulary) that are particularly relevant for speech technology, a generic typology of potential pronunciation errors is presented.
展开▼