Authoring on Demand: Natural Language Generation in Hypertext Documents

<F4.589e+05> Automatic natural language generation from underlying representations of information adds a new dimension to the electronic publishing world, particularly in the context of on-line information sources. In this paper, we provide an overview of our PebaII natural language generation system, and focus on two key themes which take on special significance in the context of hypertext document generation: the notion of co-operative discourse planning, and the generation of text that is sensitive to the context.<F5.978e+05> 1 Introduction<F4.483e+05> In this paper, we consider the use of natural language generation<F5.056e+05><F4.483e+05> (nlg) techniques in the context of dynamically constructing hypertext documents (in fact, World Wide Web pages). Our interest is in building information provision systems which can present the same information in di#erent ways at di#erent times to di#erent users, thus allowing the re-use of the underlying informational resources. A great benefi..