Corrections of disfluencies in spontaneous speech using a noisy-channel approach

In this paper we present a system which automatically cor-rects disfluencies such as repairs and restarts typically occur-ring in spontaneously spoken speech. The system is based on a noisy-channel model and its development requires no linguistic knowledge, but only manually annotated texts. Therefore, it has large potential for the rapid adaptation to new target languages. The experiments were conducted on spontaneously spoken di-alogs from the English VERBMOBIL corpus where a recall of 77.2 % and a precision of 90.2 % was obtained. To demon-strate the feasibility of rapid adaptation additional experiments on the spontaneous Mandarin Chinese CallHome corpus were performed achieving 49.4 % recall and 76.8 % precision. 1