<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-1614">
  <Title>Is it Really that Difficult to Parse German?</Title>
  <Section position="11" start_page="117" end_page="118" type="concl">
    <SectionTitle>
9 Conclusion and Future Work
</SectionTitle>
    <Paragraph position="0"> We have presented a comparative study of probabilistic treebank parsing of German, using the Negra and T&amp;quot;uBa-D/Z treebanks. Experiments with the Stanford parser, which uses a factored PCFG and dependency model, show that, contrary to previous claims for other parsers, lexicalization of PCFG models boosts parsing performance for both treebanks. The experiments also show that there is a big difference in parsing performance, when trained on the Negra and on the T&amp;quot;uBa-D/Z treebanks. This difference remains constant across  lexicalized, unlexicalized (also using the LoPar parser), and markovized models and also extends to parsing of major grammatical functions. Parser performance for the models trained on T&amp;quot;uBa-D/Z are comparable to parsing results for English with the Stanford parser, when trained on the Penn treebank. This comparison at least suggests that German is not harder to parse than its West-Germanic neighbor language English.</Paragraph>
    <Paragraph position="1"> Additional experiments with the T&amp;quot;uBa-D/Z treebank are planned in future work. A new release of the T&amp;quot;uBa-D/Z treebank has become available that includes appr. 22 000 trees, instead of the release with 15 000 sentences used for the experiments reported in this paper. This new release also contains morphological information at the POS level, including case and number. With this additional information, we expect considerable improvement in grammatical function assignment for the functions subject, accusative object, and dative object, which are marked by nominative, accusative, and dative case, respectively.</Paragraph>
  </Section>
class="xml-element"></Paper>