<?xml version="1.0" standalone="yes"?>
<Paper uid="C04-1191">
  <Title>Inferring parts of speech for lexical mappings via the Cyc KB</Title>
  <Section position="8" start_page="1" end_page="1" type="concl">
    <SectionTitle>
6 Conclusion and future work
</SectionTitle>
    <Paragraph position="0"> This paper shows that an accurate decision procedure (93.0%) accounting for the mass-count distinction can be induced from the lexical mappings in the Cyc KB. The full speech part classifier produces promising results (77.8%), considering that it is a much harder task, with over 30 categories to choose from. The features incorporate semantic information, in particular Cyc's ontological types, in addition to syntactic information (e.g., headword morphology).</Paragraph>
    <Paragraph position="1"> Future work will investigate how the classifiers can be generalized for classifying word usages in context, rather than isolated words. This could complement existing part-of-speech taggers by allowing for more detailed tag types, such as for count and agentive nouns.</Paragraph>
    <Paragraph position="2"> A separate area for future work will be to apply the techniques to other languages. For example, minimal changes to the classifier setup would be required to handle Romance languages, such as Italian. The version of the classifier that just uses Cyc reference terms could be applied as is, given lexical mappings for the language. For the combined-feature classifier, we would just need to change the list of suffixes and the part-of-speech pattern templates (from Figure 3).</Paragraph>
  </Section>
class="xml-element"></Paper>