<?xml version="1.0" standalone="yes"?>
<Paper uid="P04-1032">
  <Title>Minimal Recursion Semantics as Dominance Constraints: Translation, Evaluation, and Analysis</Title>
  <Section position="8" start_page="4" end_page="4" type="concl">
    <SectionTitle>
7 Conclusion
</SectionTitle>
    <Paragraph position="0"> We developed Niehren and Thater's (2003) theoretical translation into a practical system for translating MRS into dominance constraints, applied it systematically to MRSs produced by English Resource Grammar for the Redwoods treebank, and evaluated the results. We showed that: 1. most &amp;quot;real life&amp;quot; MRS expressions are MRSnets, which means that the translation is correct in these cases; 2. for nets, merging is not necessary (or even possible); null 3. the practical translation works perfectly for all MRS-nets from the corpus; in particular, the = q relation can be taken as synonymous with dominance in practice.</Paragraph>
    <Paragraph position="1"> Because the translation works so well in practice, we were able to compare the runtimes of MRS and dominance constraint solvers on the same inputs. This evaluation shows that the dominance constraint solver outperforms the MRS solver and displays more predictable runtimes. A researcher working with MRS can now solve MRS nets using the efficient dominance constraint solvers.</Paragraph>
    <Paragraph position="2"> A small but significant number of the MRS constraints derived by the ERG are not nets. We have argued that these constraints seem to be systematically incomplete, and their correct completions are indeed nets. A more detailed evaluation is an important task for future research, but if our &amp;quot;net hypothesis&amp;quot; is true, a system that tests whether all outputs of a grammar are nets (or a formal &amp;quot;safety criterion&amp;quot; that would prove this theoretically) could be a useful tool for developing and debugging grammars. From a more abstract point of view, our evaluation contributes to the fundamental question of what expressive power an underspecification formalism needs. It turned out that the distinction between qeq  and dominance hardly plays a role in practice. If the net hypothesis is true, it also follows that merging is not necessary because EP-conjunctions can be converted into ordinary conjunctions. More research along these lines could help unify different under-specification formalisms and the resources that are available for them.</Paragraph>
    <Paragraph position="3"> Acknowledgments We are grateful to Ann Copestake for many fruitful discussions, and to our reviewers for helpful comments.</Paragraph>
  </Section>
class="xml-element"></Paper>