CogentHelp: NLG meets SE in a tool for authoring dynamically generated on-line help

CogentHelp: NLG meets SE in a tool for authoring dynamically generated on-line help Michael White and David E. Caldwell CoGenTex, Inc. 840 Hanshaw Road Ithaca, NY 14850, USA (mike, ted}~cogentex, com Abstract CogentHelp is a prototype tool for au- thoring dynamically generated on-line help for applications with graphical user inter- faces, embodying the "evolution-friendly" properties of tools in the literate program- ming tradition. In this paper, we describe CogentHelp, highlighting the usefulness of certain natural language generation tech- niques in supporting software-engineering goals for help authoring tools -- princi- pally, quality and evolvability of help texts. 1 Introduction CogentHelp is a prototype tool for authoring dynam- ically generated on-line help for applications with graphical user interfaces (GUIs). In this paper, we describe CogentHelp, highlighting the usefulness of certain natural language generation (NLG) tech- niques in supporting software-engineering (SE) goals for help authoring tools -- principally, quality and evolvability of help texts. To our knowledge, CogentHelp is unique in that it is the first operational prototype to embody the "evolution-friendly" properties of tools in the lit- erate programming tradition (Knuth, 1992) -- e.g., the by now well-known j avadoc utility for generat- ing API documentation from comments embedded in Java source code (Friendly, 1995; cf. also John- son and Erdem, 1995; Priestly et al., 1996; Korgen, 1996) -- in a tool for generating end user-level doc- umentation. CogentHelp is also unusual in that it is (to date) one of the few tools to bring NLG tech- niques to bear on the problem of authoring dy- namically generated documents (cf. Paris and Van- der Linden, 1996; Knott et al., 1996; Hirst and Di- Marco, 1995); traditionally, most applied NLG sys- tems have focused on niches where texts can be gen- erated fully automatically, such as routine reports of various types (e.g. Goldberg et al., 1994; Kukich et al., 1994) or explanations of expert system reasoning (cf. Moore, 1995 and references therein). While striving to design highly sophisticated, fully automatic systems has undoubtedly led to a deeper understanding of the text generation process, it has had the unfortunate effect (to date) of limiting the use of techniques pioneered in the NLG community to just a few niches where high knowledge acqui- sition costs stand a chance of being balanced by substantial volume of needed texts (cf. Reiter and Mellish, 1993). By joining the emerging authoring support crowd and endeavoring to create new op- portunities in automated documentation, we hope to contribute to the broader acceptance and visi- bility of NLG technology in the overall computing community. The rest of the paper is organized as follows. In Section 2 we discuss the software engineering goals for CogentHelp. In Section 3 we provide back- ground on automated documentation and identify where CogentHelp fits in this picture. In Section 4 we give a brief overview of the CogentHelp system. In Section 5, we highlight the NLG techniques used in support of the software engineering goals identi- fied in Section 2. In Section 6 we describe Cogent- …

Content truncated to 500 words