<?xml version="1.0" standalone="yes"?>
<Paper uid="W97-0115">
  <Title>Statistical Acquisition of Terminology Dictionary*</Title>
  <Section position="6" start_page="389" end_page="389" type="concl">
    <SectionTitle>
6. Conclusion
</SectionTitle>
    <Paragraph position="0"> This research presents a chi-square method based approach to semi-automatically generate terminology dictionaries. This approach integrates such methods as new word collecting, terminology word extraction and terminology phrase generation. It significantly reduces most of the hard work which should be done manually, and reduce the effort and time which are needed to transport a natural language processing work from one domain to another. Using this terminology dictionary, encouraging results has been achieved about the coverage of terminologies.</Paragraph>
    <Paragraph position="1"> This research has practical importance in many domain related natural language applications. It can improve indexing results. It can help to decide texts' category. It also can help to rank documents with user queries. In fact, this approach will soon be embedded into an integrated Chinese information processing system - FDASCT \[ 12\] .</Paragraph>
    <Paragraph position="2"> Our future work mainly includes the utilization of deeper text processing techniques such as part of speech tagging and partial syntactic analysis in phrase generation. Word pairs would be discarded if there are no consistent syntactic relations between constituent words. And those non-noun phrases would also be discarded since terminologies are always nouns. Thus manual effort can be further reduced.</Paragraph>
  </Section>
class="xml-element"></Paper>