Basically, the project involves implementing a front end that does text manipulation for a well-established speech synthesis engine. Our goal is to improve the quality of intonation in text-to-speech synthesis through a more sophisticated treatment of discourse level factors. The overall project is sponsored by the CISE division at NSF and a description can be found as the second entry under research projects on my web page. The synthesizer that we are using is described at http://www.eloq.com. At this point we want something that would allow researchers to interactively explore different principles for implementing discourse-based intonation. Eventually we would like to work on topics like optimizing speech synthesis on web pages. Harry made a good start on the project but there is lots more to do. There really is not much math involved at all. At the most we will want some non-linear interpolation of synthesis features. I can write the equations if necessary.