<?xml version="1.0" standalone="yes"?>
<Paper uid="X98-1023">
  <Title>Improving Robust Domain Independent Summarization</Title>
  <Section position="7" start_page="176" end_page="176" type="concl">
    <SectionTitle>
4 Conclusions
</SectionTitle>
    <Paragraph position="0"> We feel that further research is warranted on improving summarization based on sentence selection and that its bad press is largely apocryphal and unjustified. In fact from a document analysts point of view material from the original document may be preferable, carrying as it does, the style and tone of the original document.</Paragraph>
    <Paragraph position="1"> We see significant opportunities in carrying out further research to develop and integrate language processing and other intelligent techniques such as those described above. One particularly challenging type of document is the HTML pages found on the web. Here techniques to identify coherent sections of text are required as well as methods for summarizing tables and groups of frames.</Paragraph>
  </Section>
class="xml-element"></Paper>