<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-1623">
  <Title>Inducing Temporal Graphs</Title>
  <Section position="11" start_page="195" end_page="195" type="concl">
    <SectionTitle>
8 Conclusions
</SectionTitle>
    <Paragraph position="0"> This paper introduces a new method for temporal ordering. The unit of our analysis is a temporal segment, a fragment of text that maintains temporal coherence. After investigating several inference strategies, we concluded that integer linear programming and best first greedy approach are valuable alternatives for TDAG construction.</Paragraph>
    <Paragraph position="1"> In the future, we will explore a richer set of constraints on the topology on the ordering graph. We will build on the existing formal framework (Fikes et al., 2003) for the verification of ordering consistency. We are also interested in expanding our framework for global inference to other temporal annotation schemes. Given a richer set of temporal relations, the benefits from global inference can be even more significant.</Paragraph>
    <Paragraph position="2">  tion over manually computed segments.</Paragraph>
  </Section>
class="xml-element"></Paper>