<?xml version="1.0" standalone="yes"?>
<Paper uid="C04-1027">
  <Title>Learning theories from text</Title>
  <Section position="9" start_page="0" end_page="0" type="concl">
    <SectionTitle>
9 Conclusions &amp; Future Work
</SectionTitle>
    <Paragraph position="0"> We have shown that it is possible to induce logically structured inference rules from parsed text. We have also shown that by using FSA techniques it is possible to construct a weighted automaton for the representation of rules/patterns generated via a knowledge mining process. This enables merging together permutations of the same pattern and facilitates human evaluation of the pattern. Furthermore, the fact that we have learned what is in effect a simple probabilistic graphical model means that we can now produce representations of this knowledge suitable for more robust inference methods of the type that we can deploy to aid reasoning and disambiguation tasks.</Paragraph>
  </Section>
class="xml-element"></Paper>