<?xml version="1.0" standalone="yes"?>
<Paper uid="W00-1203">
  <Title>Knowledge Extraction for Identification of Chinese Organization Names</Title>
  <Section position="7" start_page="19" end_page="19" type="concl">
    <SectionTitle>
5. CONCLUSIONS
</SectionTitle>
    <Paragraph position="0"> The knowledge extraction process will be continuously carried on in the future. The accumulated knowledge will be utilized for the on-line unknown word identification as well as for the off-line knowledge extraction. The proposed knowledge extraction processing model can be generalized to extract other types of linguistics or morphological knowledge, for instances, to extract the transliterate foreign names, to extract the titles of people.</Paragraph>
    <Paragraph position="1"> Some of the errors are caused by that the titles of the people are wrongly identified as organization types, since the patterns of people's name followed by their title are commonly occurred in real text. These patterns are similar to the sU'uctures of organization names. Such kind of errors can be avoid, if the titles of people are known and in fact the titles of people can be extracted by the same extraction model except that most of people's names have three characters instead of two.</Paragraph>
    <Paragraph position="2"> In the future, the knowledge extraction processes will be automatically carried out. We expect that it will be one of the major building blocks for automatic learning systems for Chinese morphology and sentence processing.</Paragraph>
  </Section>
class="xml-element"></Paper>