<?xml version="1.0" standalone="yes"?>
<Paper uid="C02-1104">
  <Title>From Shallow to Deep Parsing Using Constraint Satisfaction</Title>
  <Section position="5" start_page="0" end_page="0" type="concl">
    <SectionTitle>
5 Conclusion
</SectionTitle>
    <Paragraph position="0"> The experiments presented in this paper show that it is possible to calculate efficiently the different kind of syntactic structures of a sentence using the same linguistic resources.</Paragraph>
    <Paragraph position="1"> Moreover, the constraint-based framework proposed here makes it possible to choose the granularity, from a rough boundary detection to a deep non-deterministic analysis, via a shallow and deterministic one. The possibility of selecting a granularity level according to the data to be parsed or to the targetted application is then very useful.</Paragraph>
    <Paragraph position="2"> An interesting result for further studies lies in the perspective of combining or multiplexing different approaches. It is for example interesting to notice that common boundaries obtained by these algorithms eliminates ill-formed and least remarkable boundaries. At the same time, it increases the size of the blocks while maintaining the linguistic information available (this remains one of the most important problems for text-to-speech systems). Finally, it allows to propose a parameterized granularity in balancing the relative importance of different competing approaches.</Paragraph>
  </Section>
class="xml-element"></Paper>