<?xml version="1.0" standalone="yes"?>
<Paper uid="H93-1072">
  <Title>AN OVERVIEW OF DR-LINK AND ITS APPROACH TO DOCUMENT FILTERING</Title>
  <Section position="11" start_page="360" end_page="360" type="concl">
    <SectionTitle>
6. CONCLUSION
</SectionTitle>
    <Paragraph position="0"> Our implementation and testings of the SFCOder as a means for semantically representing the content of texts, either for the purpose of ranking a document set according to likelihood of being relevant to an individual query or for producing conceptually related clusters of documents for browsing are very promising.</Paragraph>
    <Paragraph position="1"> Particularly worthy of note is the observation that in a large operational system, the ability to filter out an average of 72% of the incoming flux of millions of documents will have a significant impact on any document detection system's performance with which this semantic-based document falter is combined.</Paragraph>
  </Section>
class="xml-element"></Paper>