<?xml version="1.0" standalone="yes"?>
<Paper uid="W01-0720">
  <Title>A Psychologically Plausible and Computationally Effective Approach to Learning Syntax</Title>
  <Section position="8" start_page="0" end_page="0" type="concl">
    <SectionTitle>
7 Conclusions
</SectionTitle>
    <Paragraph position="0"> There is further work to be completed in extending the system to allow it to deal with movement and thus the whole of the Penn Treebank. Further investigation of parameters of CLL should also be completed. Further work needs to be done in building an accurate gold standard corpus. There is also a possibility of performing experiments on sequences of parts-of-speech, as Joshi and Srinivas (Joshi and Srinivas, 1994) and Osborne and Briscoe (Osborne and Briscoe, 1997) did. This would reduce the effects of the sparse data problem. null However, we have presented a system that is psychologically plausible and whose results show that, given the complexity of the problem attempted, it is computationally effective. The results compare reasonably with systems attempting much simpler and psychologically less plausible problems.</Paragraph>
  </Section>
class="xml-element"></Paper>