<?xml version="1.0" standalone="yes"?>
<Paper uid="E99-1002">
  <Title>Generating referring expressions with a unification grammar</Title>
  <Section position="8" start_page="12" end_page="13" type="concl">
    <SectionTitle>
4 Conclusion
</SectionTitle>
    <Paragraph position="0"> Two ideas have been suggested: The linguistic context relevant to choosing nominal referring expressions can be formalized, in part, by vectors giving focus values and mention ratios for all potential referents.</Paragraph>
    <Paragraph position="1"> These features can be threaded through the text structure during generation by assigning initial and final contexts to each textual unit. * Since generation requires search through a space of possible structures, there is a danger that expensive computations of linguistic context will be repeated many times.</Paragraph>
    <Paragraph position="2"> This can be avoided by composing 'bespoke' phrase-structure rules, tailored to the entities currently in the knowledge base, before embarking on the search phase.</Paragraph>
    <Paragraph position="3"> Note that the first proposal can be employed independently from the second, which is more speculative. However, we think that the idea of using specially tailored phrase-structure rules deserves consideration. Its applications are not limited to the generation of referring expressions. One aim of the ICONOCLAST project is to generate texts in a variety of house styles, where a 'style' embraces preferences regarding textual organization, wording, punctuation and layout. To cover a large range of styles, many patterns must be made available to the generator, even though only a fraction are relevant for a particular company and a particular knowledge base. Before commencing a search through this space of patterns, it is worth devoting some effort to refining the search space by filtering out irrelevant rules and perhaps merging rules that separately constrain linguistic and presentational features.</Paragraph>
    <Paragraph position="4"> The efficiency of the approach suggested here is difficult to evaluate in general terms: it will depend on the nature of the alternative methods, and also on the size of the generated text. For larger texts, in which entities may be mentioned many times, the initial investment of effort in creating bespoke phrase-structure rules will obviously pay more dividends. However, before trying to evaluate this difficult trade-off, we feel the next step should be to ensure that the approach can be applied to a wider range of referring expressions (e.g. demonstratives, plurals), and that it can be extended to cover a more complex treatment of focus such as centering theory (Walker et al., 1998).</Paragraph>
    <Paragraph position="5"> Although we have not addressed here the problem of selecting appropriate properties for use in  Proceedings of EACL '99 referential descriptions (Dale and Reiter, 1995), it is worth noting that since this selection depends on the current state of the knowledge base, it can also be performed before the search phase of generation, the results of the selection algorithm being saved in the form of additional bespoke rules.</Paragraph>
  </Section>
class="xml-element"></Paper>