<?xml version="1.0" standalone="yes"?>
<Paper uid="P06-2028">
  <Title>Using Lexical Dependency and Ontological Knowledge to Improve a Detailed Syntactic and Semantic Tagger of English</Title>
  <Section position="9" start_page="220" end_page="221" type="concl">
    <SectionTitle>
7 Conclusion
</SectionTitle>
    <Paragraph position="0"> We have described a method for simultaneously labeling the syntax and semantics of words in running text. We develop this method starting from a state-of-the-art maximum entropy POS tagger which itself outperforms previous attempts to tag this data (Black et al., 1996b). We augment this tagging model with two distinct types of knowledge: the identity of dependent words in the sentence, and word class membership information of the word being tagged. We define the features in such a manner that the useful lexical dependencies are selected by the model, as is the granularity of the word classes used. Our experimental results show that large gains in performance are obtained using each of the techniques. The dependent words boosted overall performance, especially when tagging verbs. The hierarchical ontology-based approaches also increased over-all performance, but with particular emphasis on OOV's, the intended target for this feature set.</Paragraph>
    <Paragraph position="1"> Moreover, when features from both knowledge sources were applied in combination, the gains were cumulative, indicating little overlap.</Paragraph>
    <Paragraph position="2"> Visual inspection the output of the tagger on held-out data suggests there are many remaining errors arising from special cases that might be better handled by models separate from the main tagging model. In particular, numerical expressions  andnamedentitiescauseOOVerrorsthatthetechniquespresentedinthispaperareunabletohandle. null In future work we would like to address these issues, and also evaluate our system when used as a componentofaWSDsystem, andwhenintegrated within a machine translation system.</Paragraph>
    <Paragraph position="3">  interval of the mean at a 95% significance level, calculated using bootstrap resampling.</Paragraph>
  </Section>
class="xml-element"></Paper>