<?xml version="1.0" standalone="yes"?>
<Paper uid="W04-2407">
  <Title>Memory-Based Dependency Parsing</Title>
  <Section position="6" start_page="0" end_page="0" type="concl">
    <SectionTitle>
5 Conclusion
</SectionTitle>
    <Paragraph position="0"> In this paper we have shown that a combination of memory-based learning and deterministic dependency parsing can be used to construct a robust and efficient parser for unrestricted natural language text, achieving a parsing accuracy which is close to the state of the art even with relatively limited amounts of training data. Classifiers based on memory-based learning achieve higher parsing accuracy than previous probabilistic models, and the improvement increases if lexical information is added to the model.</Paragraph>
    <Paragraph position="1"> Suggestions for further research includes the further exploration of alternative models and parameter settings, but also the combination of inductive and analytical learning to impose high-level linguistic constraints, and the development of new parsing methods (e.g. involving multiple passes over the data). In addition, it is important to evaluate the approach with respect to other languages and corpora in order to increase the comparability with other approaches.</Paragraph>
  </Section>
class="xml-element"></Paper>