<?xml version="1.0" standalone="yes"?>
<Paper uid="P99-1065">
  <Title>A Statistical Parser for Czech*</Title>
  <Section position="6" start_page="511" end_page="511" type="concl">
    <SectionTitle>
6 Conclusions
</SectionTitle>
    <Paragraph position="0"> The 80% dependency accuracy of the parser represents good progress towards English parsing performance. A major area for future work is likely to be an improved treatment of morphology; a natural approach to this problem is to consider more carefully how POS tags are used as word classes by the model. We have begun to investigate this issue, through the automatic derivation of POS tags through clustering or &amp;quot;splitting&amp;quot; approaches. It might also be possible to exploit the internal structure of the POS tags, for example through incremental prediction of the POS tag being generated; or to exploit the use of word lemmas, effectively splitting word-word relations into syntactic dependencies (POS tag-POS tag relations) and more semantic (lemma-lemma) dependencies.</Paragraph>
  </Section>
class="xml-element"></Paper>