<?xml version="1.0" standalone="yes"?>
<Paper uid="I05-2020">
  <Title>Effect of Domain-Specific Corpus in Compositional Translation Estimation for Technical Terms</Title>
  <Section position="5" start_page="116" end_page="118" type="evalu">
    <SectionTitle>
5 Experiments and Evaluation
</SectionTitle>
    <Paragraph position="0"/>
    <Section position="1" start_page="116" end_page="117" type="sub_section">
      <SectionTitle>
5.1 Translation Pairs for Evaluation
</SectionTitle>
      <Paragraph position="0"> In our experimental evaluation, within the framework of compiling a bilingual lexicon for technical terms, we evaluate the translation estimation part which is indicated with bold line in Fig- null , the rate of including correct translations within the collected domain/topic specific corpus ure 4. In the evaluation of this paper, we simply skip the evaluation of the process of collecting technical terms to be listed as the headwords of a bilingual lexicon. In order to evaluate the translation estimation part, from ten categories of existing Japanese-English technical term dictionaries listed in Table 2, terms are randomly picked up for each of the set X  , the rate of including correct translation within the collected domain/topic specific corpus, respectively.</Paragraph>
    </Section>
    <Section position="2" start_page="117" end_page="117" type="sub_section">
      <SectionTitle>
5.2 Translation Selection from Existing
Bilingual Lexicon
</SectionTitle>
      <Paragraph position="0"> For the terms of X</Paragraph>
      <Paragraph position="2"> , the selected translations are judged by a human. The correct rates are 69% from English to Japanese on the average and 75% from Japanese to English on the average.</Paragraph>
    </Section>
    <Section position="3" start_page="117" end_page="117" type="sub_section">
      <SectionTitle>
5.3 Compositional Translation Estimation
for Technical Terms without the
Domain/Topic Specific Corpus
</SectionTitle>
      <Paragraph position="0"> Without the domain specific corpus, the correct rate of the first ranked translation candidate is 19% on the average (both from English to Japanese and from Japanese to English). The rate of including correct candidate within top 10 is 40% from English to Japanese and 43% from Japanese to English on the average. The rate of compositionally generating correct translation using both Eijiro and the bilingual constituents lexicons (n = [?]) is about 50% on the average (both from English to Japanese and from Japanese to English).</Paragraph>
    </Section>
    <Section position="4" start_page="117" end_page="118" type="sub_section">
      <SectionTitle>
5.4 Compositional Translation Estimation
</SectionTitle>
      <Paragraph position="0"> for Technical Terms with the  With domain specific corpus, on the average, the correct rate of the first ranked translation candidate improved by 8% from English to Japanese and by 2% from Japanese to English. However, the rate of including correct candidate within top 10 decreased by 7% from English to Japanese, and by 14% from Japanese to English. This is because correct translation does not exist in the corpus for 32% (from English to Japanese) or 43% (from Japanese to English) of the 667 translation pairs for evaluation.</Paragraph>
      <Paragraph position="1"> For about 35% (from English to Japanese) or 30% (from Japanese to English) of the 667 translation pairs for evaluation, correct translation does exist in the corpus and can be generated through the compositional translation estimation process. For those 35% or 30% translation pairs, Figure 5 compares the correct rate of the first ranked translation pairs between with/without the domain/topic specific corpus. The correct rates increase by 34[?]37% with the domain/topic specific corpus. This result supports the claim that the do- null and can be Generated Compositionally main/topic specific corpus is effective in translation estimation of technical terms.</Paragraph>
    </Section>
  </Section>
class="xml-element"></Paper>