<?xml version="1.0" standalone="yes"?>
<Paper uid="H01-1043">
  <Title>Japanese Case Frame Construction by Coupling the Verb and its Closest Case Component</Title>
  <Section position="7" start_page="5" end_page="5" type="evalu">
    <SectionTitle>
7. EXPERIMENTS AND DISCUSSION
</SectionTitle>
    <Paragraph position="0"> Since it is hard to evaluate the dictionary statically, we usethedictionaryincasestructureanalysisandevaluatethe analysis result. We used 200sentences ofMainichiNewspaperCorpusasatestset. Weanalyzed casestructures ofthe sentences using the method proposed by [4]. Asthe evaluationofthecasestructureanalysis,wecheckedwhethercases null of ambiguous case components (topic markers and clausal modifiers) are correctly detected or not. The evaluation result is presented in Table 2. The baseline is the result by assigningavacantcaseinorderof'ga','wo',and'ni'. When we do not consider parsing errors to evaluate the case detection, the accuracy of our method for topic markers was 96% and that for clausal modifiers was 76%. The baseline accuracy for topic markers was 91% and that for clausal modifiers was 62%. Thus we see our method is superior to the baseline.</Paragraph>
    <Paragraph position="1">  first is clausal modifiers which have no case relation to the modifeessuchas&amp;quot;*** wo mitomeru houshin&amp;quot;'policyofconsenting ***'(+ above). The Second is verbs which take two ga 'nominative' case markers (one is wa superficially) such as &amp;quot;gyokai wa *** toiu tokutyo ga aru&amp;quot; 'industries have the characteristic of ***'(++ above). Handling these phenomena is an area of future work.</Paragraph>
  </Section>
class="xml-element"></Paper>