<?xml version="1.0" standalone="yes"?>
<Paper uid="A97-2006">
  <Title>An Improvement in the Selection Process of Machine Translation Using Inductive Learning with Genetic Algorithms</Title>
  <Section position="4" start_page="0" end_page="0" type="metho">
    <SectionTitle>
2 Outline of Translation Method
</SectionTitle>
    <Paragraph position="0"> Figure 1 shows the outline of our proposed translation method. First, the user inputs a source sentence in English. Second , in the translation process, the system produces several candidates of translation results using translation rules extracted in the learning process. Third, the user proofreads the translated sentences if they include some errors. Fourth, in the feedback process, the system determines the fitness value of translation rules used in the translation process and performs the selection process of erroneous translation rules. In the learning process, new translation examples are automatically produced by crossover and mutation, and various translation rules are extracted from the translation examples by inductive learning.</Paragraph>
  </Section>
  <Section position="5" start_page="0" end_page="11" type="metho">
    <SectionTitle>
3 Improvement in Selection Process
</SectionTitle>
    <Paragraph position="0"> In the previous method of selection process described in Section 2, translation rules are evaluated only when they are used in the translation process.</Paragraph>
    <Paragraph position="1"> These translation rules are part of all the translation rules in the dictionary. Therefore, many erroneous  To resolve this problem, we propose an improvement in the selection process. Our proposed improvement does not require any analytical knowledge as initial condition. Methods that use analytical knowledge have some problems, such as difficulty in dealing with unregistered words. We consider that this problem can be resolved by the learning method without any analytical knowledge. Therefore, we consider that our proposed improvement can remove many erroneous translation rules by utilizing only the given translation examples without the requirement of analytical knowledge.</Paragraph>
    <Paragraph position="2"> The system evaluates the translation rules by utilizing the given translation examples directly.</Paragraph>
    <Paragraph position="3"> Namely, it determines whether a combination of the English word and the Japanese word in a translation rule is true or false by utilizing the given translation examples. The combination may be true when it exists in a given translation example. For example, the combination of words which are &amp;quot;I&amp;quot; in English and &amp;quot; Watashi 1 (In Japanese &amp;quot;I&amp;quot;)&amp;quot; in Japanese is true when this combination exists in a given translation example. On the other hand, the combination of words which are &amp;quot;volleyball&amp;quot; in English and &amp;quot;Basukettoboru(In Japanese &amp;quot;basketball&amp;quot;)&amp;quot; in Japanese is false when this combination does not exist in all given translation examples. In the all combinations of words in a translation rule, the system determines whether the each combination of words is true or false. And the system determines the rate of error based on the number of erroneous combinations, and removes the translation rules for which the rate of error is high.</Paragraph>
  </Section>
  <Section position="6" start_page="11" end_page="11" type="metho">
    <SectionTitle>
4 Experiments
</SectionTitle>
    <Paragraph position="0"> In the experiments, 461 translation examples were used as data. The examples were taken from a textbook (Hasegawa et al., 1991) for first-grade junior 1Italic means pronunciation of Japanese high school students. All of the translation examples were processed by the method outlined in Figure 1. The initial dictionary was empty. The experiments were carried out with and without the improvement for the selection process described in Section 3. In the experiments, the precision increased from 87.5% to 93.7% and the recall increased from 4.5% to 56.0%.</Paragraph>
  </Section>
class="xml-element"></Paper>