<?xml version="1.0" standalone="yes"?>
<Paper uid="C00-2101">
  <Title>Learning Semantic-Level Information Extraction Rules by Type-Oriented ILP</Title>
  <Section position="7" start_page="702" end_page="702" type="evalu">
    <SectionTitle>
6.2 Results
</SectionTitle>
    <Paragraph position="0"> 'l?M)le 4 shows the results of our experiment. In the experiment of learning from semantic representations, including errors in case-role selection and semantic category selection, precision was 3We used ~rticles from the Mainichi Newspaimrs of 1994 with permission.</Paragraph>
    <Paragraph position="1"> very high. 'l'he precision of the learned rules lot price was low beta.use the seman tic category name automatieaJly given to the price expressions in the dat~ were not quite a.ppropriate. For the tire items, 6?-82% recall was achieved.</Paragraph>
    <Paragraph position="2"> With the background knowledge having sere antic representations corrected by hand, precision was very high mid 70-88% recMl was achieved.</Paragraph>
    <Paragraph position="3"> The precision of price was markedly improved.</Paragraph>
    <Paragraph position="4"> It ix important that the extraction of live ditthrent pieces o1' information showed good results. This indica.tex that the \]LI' system RIII~ + has a high potential in IE tasks.</Paragraph>
  </Section>
class="xml-element"></Paper>