<?xml version="1.0" standalone="yes"?>
<Paper uid="W04-1011">
  <Title>Handling Figures in Document Summarization</Title>
  <Section position="9" start_page="0" end_page="0" type="relat">
    <SectionTitle>
8 Related Work
</SectionTitle>
    <Paragraph position="0"> Automated text summarization has advanced substantially in the last decade. See for example, the major collection of papers, (Mani &amp; Maybury, 1999) and the special journal issue (Radev, Hovy, &amp; McKeown, 2002). Reviews include (Hovy, 2002; Marcu, 2003). A recent useful monograph is (Mani, 2001). Another recent work is (Barzilay, 2003), focused on mult idocument summarization and going beyond sentence extraction to consider phrases.</Paragraph>
    <Paragraph position="1"> Paradoxically, work on the summarization of scientific articles is inhibited by the fact that virtually all scientific articles have abstracts as a sta ndard component. But there are other tasks such as developing user-tailored summaries (Teufel &amp; Moens, 2002).</Paragraph>
    <Paragraph position="2"> The generation of coordinated explanations involving text and graphics offers insight into the relations between them (Feiner &amp; McKeown, 1990). This task involves dealing with the internal structure of diagrams, as do problems of image retrieval, which can be aided by developing ontology-based descriptions of the images (Hyvonen, Styrman, &amp; Saarela, 2002).</Paragraph>
    <Paragraph position="3"> Diagrams form a part of a coordinated discourse, so that diagram summarization can profit from the work done on text summarization that focuses on discourse structure. Examples of discourse-related approaches include (Boguraev &amp; Neff, 2000; Marcu, 1997a, 1997b; Teufel &amp; Moens, 2002).</Paragraph>
  </Section>
class="xml-element"></Paper>