<?xml version="1.0" standalone="yes"?>
<Paper uid="W99-0105">
  <Title>O @ O O O 0 O O @ O O @ O 0 O O @ Positing and resolving bridging anaphora in deverbal NPs</Title>
  <Section position="5" start_page="43" end_page="44" type="concl">
    <SectionTitle>
4 Conclusion and Future work
</SectionTitle>
    <Paragraph position="0"> This paper presents an extension of computattional approa~ee to bridging in deverbal NPs that use the verb argument structure. Unlike other works on nominalisations which aim at a full description of the status of each argument in the deverbal noun, our approach is mainly concerned with the coding of the 'minimai' information which is necessary for safely positing and resolving bridging phenomena.</Paragraph>
    <Paragraph position="1"> Minimal de~dptions are best suited for our computational task because they trigger the hypothesis that there is s bridging in a reduced number of cases, with a higher degree of certainty and provide precise information for its resolution.</Paragraph>
    <Paragraph position="2"> For a deeper evaluation of the effectiveness of the heuristics presented here, we plan to extend our texture resolution algorithm according to the computational rules described in section 3, in order to run verification tests on &amp; larger corpus. In this __-~co__nd phase, additional effort will be invested for the annotation of separate training and test corpus which will be used for a more objective evaluation of precision and recall. Another issue we would like to investigate is the extension of our notion of/ueour/te to other cases of nominal roles, for example meronyms and attributes. This would not require additional annotation in the lexicon, given that in a conceptual dictionary like WordNet meronyms are already present and there are plans to insert attributes (Fellbaum, 1998).</Paragraph>
  </Section>
class="xml-element"></Paper>