This paper describes some of the linguistic research underlying the development of a speech-driven interface, and outlines the basic steps in data capture and analysis which are required to produce the model of linguistic activity in the office context needed to ensure full functionality. Particular attention is paid to the results of some of our linguistic modelling from dictation data. Keywords: speech recognition and understanding, natural language interfaces, office automation, spoken dialogue.
Bibliographic reference. Monaghan, James / Cheepen, Christine (1991): "Linguistic modelling for a speech interface in the office context", In EUROSPEECH-1991, 745-748.