<?xml version="1.0" standalone="yes"?>
<Paper uid="C02-2013">
  <Title>SOAT: A Semi-Automatic Domain Ontology Acquisition Tool from Chinese Corpus</Title>
  <Section position="6" start_page="0" end_page="0" type="evalu">
    <SectionTitle>
4.2 Results from Different Corpora
</SectionTitle>
    <Paragraph position="0"> We select three different corpora from different information resources in the &amp;quot;network&amp;quot; domain.</Paragraph>
    <Paragraph position="1"> The first corpus is a collection of FAQ sentences about computer network. The second corpus is a collection of sentences containing the keyword &amp;quot;network&amp;quot; from the CKIP corpus. The third corpus is the collection of sentences from Windows 2000 online help document. To reduce the cost of human verification, we limit the size of corpus to 275 sentences. The result in Table 6 shows that there is a trade-off between extraction rate and the accuracy rate.</Paragraph>
    <Paragraph position="2"> Table 6. The extraction and accuracy rate of three corpora in the same domain</Paragraph>
    <Section position="1" start_page="0" end_page="0" type="sub_section">
      <SectionTitle>
4.3 The Advantage of a Semi-Automatic
</SectionTitle>
      <Paragraph position="0"> Domain Ontology Editor for QA System SOAT can help in QA system ontology editing.</Paragraph>
      <Paragraph position="1"> In our experience, a trained knowledgeable editor can compile about 100 FAQs into our ontology manually per day. On the other hand, with the help of SOAT, a knowledgeable editor can edit on the average 4 categories, 25 attributes and 42 activities that SOAT extracted. The quantity is estimated on 4*(25+42)=268 FAQ query concepts at least. Thus, the productivity of using SOAT is approximated 268% times. It is obvious that SOAT can help reducing the cost of building a new domain ontology.</Paragraph>
    </Section>
  </Section>
class="xml-element"></Paper>