<?xml version="1.0" standalone="yes"?>
<Paper uid="C90-3085">
  <Title>Lexicon I Template Filter Words ~1 ndexes Textual Stories \[Parser i I I %c~Ttl to.co, i Structures ~ I~ Terms An Architectural Overview of NAS</Title>
  <Section position="2" start_page="0" end_page="0" type="intro">
    <SectionTitle>
SCOPE AND OB,IECTIVES
</SectionTitle>
    <Paragraph position="0"> A primary design goal behind NAS is to develop a system employing NLP technology that wouh:t be capable of either routing news t'rom electronic news feeds in real-time or indexing news with respect to very large sets of indexing terms (authority files). ~ This vocabulary is broad and ranges over diverse domains, with lerms representing proper names, concrete objects, or abslract relationships. The last type is dependent on the content of the stories and is particularly suited for the syntactic/ semantic techniques of NAS.</Paragraph>
    <Paragraph position="1"> The form of the indexes that NAS produces is a set of pairs of headings and subheadings or descriptors. The headings are frequently proper names while the descriptors add detailed information to the index by denoting various relationships of entities mentioned in the news reports.</Paragraph>
  </Section>
class="xml-element"></Paper>