<?xml version="1.0" standalone="yes"?>
<Paper uid="W04-0814">
  <Title>The University of Amsterdam at Senseval-3: Semantic Roles and Logic Forms</Title>
  <Section position="6" start_page="0" end_page="0" type="concl">
    <SectionTitle>
5 Conclusions
</SectionTitle>
    <Paragraph position="0"> Our main finding during the development of the systems for the two Senseval tasks was that semantic relations are indeed very close to syntactic dependencies. Using deep dependency structures helped to keep the manual rules for the LF task simple and made the learning for the SR task easier. Also we found that memory-based learning can be efficiently applied to complex, highly structured problems such as the identification of semantic roles.</Paragraph>
    <Paragraph position="1"> Our future work includes more accurate fine-tuning of the learner for the SR task, extending the coverage of the LF generator, and experimenting with the generated LFs for question answering.</Paragraph>
  </Section>
class="xml-element"></Paper>