<?xml version="1.0" standalone="yes"?>
<Paper uid="J95-1002">
  <Title>Expressing Rhetorical Relations in Instructional Text: A Case Study of the Purpose Relation</Title>
  <Section position="3" start_page="30" end_page="31" type="relat">
    <SectionTitle>
2. Related Work in Natural Language Generation
</SectionTitle>
    <Paragraph position="0"> In pursuit of other issues, many studies have adopted a temporary solution to the problem of managing diverse forms of expression, namely, that of choosing a single lexical and grammatical form to express each of the relevant types of information dealt with by the system. It is then a simple matter of allowing the type of information to determine the appropriate expressional form. Hovy's text structurer (Hovy 1988b), for example, uses rhetorical relations as defined in Rhetorical Structure Theory (RST) (Mann and Thompson 1987) to order a set of propositions to be expressed. It does not make any decision as to how a chosen relation is to be expressed, but rather leaves this task to the rudimentary implementation of the rhetorical relations provided by the Penman text generation system (Mann 1985). In the case of a purpose, for example, Penman will produce a non-fronted in order to infinitive clause, as in the following output of the structurer (Hovy 1988b, p. 167): &amp;quot;Knox is en route in order to rendezvous with CTG 070.10.&amp;quot; A similar approach to expressing rhetorical relations was taken in McKeown's TEXT system (McKeown 1985).</Paragraph>
    <Paragraph position="1"> Text generators specifically designed for instructional text, such as Mellish and Evans' generator (Mellish and Evans 1989), EPICURE (Dale 1992), COMET (McKeown et al. 1990), and TECHDOC (R6sner and Stede 1992b), display similar characteristics.</Paragraph>
    <Paragraph position="2"> (See the analytical work of Delin, Scott, and Hartley \[1993\] for a notable exception to this pattern.) Mellish and Evans' generator, for example, uses the output of NONLIN, a non-linear planner (Tate 1976), as the preliminary rhetorical structure for the text. Because this often produced text that was monotonous or hard to understand, they included what was termed a message optimization phase that specifies rules for removing or modifying certain elements of the plan structure that are known to produce poor text.</Paragraph>
    <Paragraph position="3"> Although this greatly improves the text, it still tends toward text that is difficult to read. This problem is, in part, due to the fact that some of the plans they look at are quite complex and correspondingly difficult to express, but it is also attributable to the lack of a detailed corpus study of the linguistic tools used by technical writers in instructional text. The IMAGENE project can be seen as an extension of their work  Computational Linguistics Volume 21, Number 1 that employs such a study to help manage diversity of forms of expression. There are other natural language generation projects that have addressed similar issues. Two such examples are Hovy's Pauline (Hovy 1988a) and Meteer's Spokesman (Meteer 1991, 1992), both of which are based, at least in part, on corpus studies.</Paragraph>
    <Paragraph position="4"> Pauline produces an impressive range of expressional forms that are based on a list of pragmatic features of the communicative environment, including information about the conversational atmosphere, the speaker, the hearer, the relationship between the two, and the interpersonal communicative goals of the speaker. Its construction required a considerable amount of analysis of sample texts, but unfortunately, very little is said about how this analysis was actually performed and how well the text produced by Pauline matches the text in the corpus. The concerns of the IMAGENE project are similar, except that both the text type (instructional text) and the linguistic phenomenon (expressing rhetorical relations) are much more focused. The results of our study are therefore more detailed, but also more constrained (see the concept of domain communication knowledge, as described by Kittredge, Korelsky, and Rambow \[1991\]).</Paragraph>
    <Paragraph position="5"> In the Spokesman project, Meteer proposed an architecture for addressing what she termed the problem of expressibility in text planning (Meteer 1992). Her fundamental thesis is that an abstract linguistic representation is needed to provide the text planner with information on constraints of expression (see Vaughan, and McDonald 1987).</Paragraph>
    <Paragraph position="6"> Her constraints are taken, at least in part, from a study of text revisions made by expert editors. The IMAGENE project concurs with this concern for detailed forms of expression, but its methodology is geared toward determining the elements of the communicative context used to choose between equally acceptable alternative forms of expression. Meteer does not address what to do if, after using her constraints to remove unacceptable forms of expression, there are a number of remaining acceptable forms. This issue of choice is central to the current study.</Paragraph>
  </Section>
class="xml-element"></Paper>