Using two-level morphology to transcribe Swedish names

Names are difficult to handle for normal letter-to-sound rules, since these usually are designed for ordinary words. The structure of Swedish names differ from ordinary words but their multi-morphemic structure make them suitable to analyse with a morphological analyser. The paper presents the work on names from the Swedish telephone directory, as part of the ONOMASTICA project [7], including a brief study of the structure of Swedish names. The speech communication group at KTH have developed a system where a morphology analyser is used together with a set of rules to transcribe ordinary Swedish words. This paper will describe the work done to extend this system to cope with names as well. The paper shows that the approach of transcribing Swedish names with the Two-level Morphology analyser (TWOL) is appropriate.