Appeal No. 1998-0650 Application 08/548,794 Appellant discloses on pages 1 through 4 of the specification, that there is a need to assign appropriate intonation to the text so that the speech generated will have a quality of naturalness. Appellant discloses that intonation includes such features as variations in prominence, pitch range, intonational contour, and intonational phrasing. Appellant discloses a method of training a text-to-speech system that involves taking a set of predetermined text and having a human annotate it with intonational feature annotations. This results in annotated text. Next, the structure of the set of predetermined text is analyzed to generate information which is a statistical representation. The statistical representation may be repeatedly used to generate synthesized speech for new sets of input text without training the text- to-speech system further. We find that the claim language recited in Appellant's claim 1 recites subject matter that is a practical application of generating statistical representation of intonational feature information for use in training a text-to-speech system. 7Page: Previous 1 2 3 4 5 6 7 8 9 10 NextLast modified: November 3, 2007