<?xml version="1.0" standalone="yes"?>
<Paper uid="P96-1058">
  <Title>Maximizing Top-down Constraints for Unification-based Systems</Title>
  <Section position="6" start_page="382" end_page="382" type="concl">
    <SectionTitle>
5 Discussion and Future Work
</SectionTitle>
    <Paragraph position="0"> The proposed method has an obvious difficulty: the complexity caused by the repeated propagations could become overwhelming for some grammars. However, in the experiment on LINK system using a fairly broad grammar (over 130 rules), precompilation terminated with only a marginally longer processing time.</Paragraph>
    <Paragraph position="1"> In the experiment, all features (around 40 syntactic/semantic features) except for one in the example in this paper were able to be used in propagation.</Paragraph>
    <Paragraph position="2"> In the preliminary analysis, the number of edges entered into the chart has decreased by 30% compared to when only the category feature (i.e., context-free backbone) was used in propagation.</Paragraph>
    <Paragraph position="3"> For future work, we intend to apply the proposed method to other grammars. By doing the empirical analysis of precompilation and parse efficiency for different grammars, we will be able to conclude the practical applicability of the proposed method.</Paragraph>
    <Paragraph position="4"> We also indend to do more exhaustive case analysis and investigate the selection ordering of the detection function. Although the current definition covers most cases, it is by no means complete.</Paragraph>
  </Section>
class="xml-element"></Paper>