<?xml version="1.0" standalone="yes"?>
<Paper uid="W99-0314">
  <Title>Automatically Extracting</Title>
  <Section position="12" start_page="113" end_page="113" type="concl">
    <SectionTitle>
6 Conclusion
</SectionTitle>
    <Paragraph position="0"> We have presented an automatic mapping from DRI backward- and forward-looking tags to grounding features and discourse units. Our approach assumes simultaneous segmentation into utterance units and annotation of BF tags, which eliminates the need to split or join utterances. The mapping is still being tested but preliminary comparison with a human annotator was * promising. Automatic derivation of grounding tags will eliminate the need for separate annotation of grounding, making dialogue annotation quicker and removing a possible source of error.</Paragraph>
  </Section>
class="xml-element"></Paper>