<?xml version="1.0" standalone="yes"?>
<Paper uid="P98-1099">
  <Title>Combining Multiple, Large-Scale Resources in a Reusable Lexicon for Natural Language Generation</Title>
  <Section position="8" start_page="612" end_page="612" type="concl">
    <SectionTitle>
4 Conclusion
</SectionTitle>
    <Paragraph position="0"> In this paper, we present research on building a rich, large-scale, and reusable lexicon for generation by combining multiple heterogeneous linguistic resources. Novel semi-automatic transformation and integration were used in combining resources to ensure reliability of the resulting lexicon. The lexicon, together with a multi-level feedback architecture, is used in a practical generation system, PlanDOC.</Paragraph>
    <Paragraph position="1"> The application of the lexicon in a generation system such as PlanDOC has many advantages.</Paragraph>
    <Paragraph position="2"> First, paraphrasing power of the system can be greatly improved due to the introduction of synonyms at the lexical concept level and alternations at the syntactic level. Second, the integration of the lexicon and the flexible architecture enables us to separate the domain-dependent component of the lexical choice module from domain-independent components so they can be reused. Third, the integration of the lexicon with the surface realization system helps in checking for grammatical errors and also simplifies the interface input to the realization system. For these reasons, we were able to develop PlanDOC system in a short time.</Paragraph>
    <Paragraph position="3"> Although the lexicon was developed for generation, it can be applied in other applications too. For example, the syntactic-semantic constraints can be used for word sense disambiguation (Jing et al., 1997); The subcategorization and alternations from EVCA/COMLEX are better resources for parsing; WordNet enriched with syntactic information might also be of value to many other applications.</Paragraph>
  </Section>
class="xml-element"></Paper>