<?xml version="1.0" standalone="yes"?>
<Paper uid="W00-0708">
  <Title>Memory-Based Learning for Article Generation</Title>
  <Section position="9" start_page="47" end_page="47" type="concl">
    <SectionTitle>
6 Concluding remarks
</SectionTitle>
    <Paragraph position="0"> We described a memory-based approach to automated article generation that uses a variety of lexical, syntactic and semantic features as provided by the Penn Treebank Wall Street Journal data and a large hand-encoded MT dictionary. With this approach we achieve an accuracy of 82.6%. We believe that this approach is an encouraging first step towards a statistical device for automated article generation that can be used in a range of applications such as speech prosthesis, machine translation and automated summarization.</Paragraph>
  </Section>
class="xml-element"></Paper>