The Use and Potential of Extensible Mark-up (xml) in Speech Generation

discusses the importance of the extensible markup language (XML) to the development of current and future speech synthesis systems. XML is an emerging standard of textual markup which is well suited to efficient computer manipulation. It provides a means for representing and processing the complex linguistic structures used in synthesis in an open and non-proprietary manner. Currently XML is being used to markup text for input into text-to-speech systems, for the annotation of corpora, and for the implementation of speech driven applications. However recent research has shown how XML can be used to create an open architecture for synthesis through the markup of working data structures and through the creation of a language for manipulating XML data. The article concludes by suggesting that work towards standards for marking up the information structure of text and its discourse function with XML would help create a new generation of intelligent sounding computer voices.