<?xml version="1.0" standalone="yes"?>
<Paper uid="W04-1205">
  <Title>Zone Identification in Biology Articles as a Basis for Information Extraction</Title>
  <Section position="9" start_page="34" end_page="34" type="concl">
    <SectionTitle>
7 Conclusion
</SectionTitle>
    <Paragraph position="0"> We have provided a qualitative analysis of the process and results of ZI based on our hand-annotated sample, with a view to strengthening the basis for the annotation scheme. We are now starting to use our sample as training data for machine learning, as well as creating more data in a systematic way, toward automatic annotation.</Paragraph>
    <Paragraph position="1"> We are also considering to use our ontology management tool (Open Ontology Forge, http://research.nii.ac.jp/~collier/resources/OOF/ind ex.htm) for these purposes; 1) to define zone classes as ontology classes; zone annotation is then expected to be a variant of named entity annotation, which we are familiar with, and 2) to link between expressions referring to results (e.g. these results/ our results) and their antecedent (i.e. the RSL zone providing a concrete description of the experimental results), using the coreference tool.</Paragraph>
    <Paragraph position="2"> Applications include full color representation of annotated texts; a sample is available at: http://research.nii.ac.jp/~collier/projects/ZAISA/in dex.htm. Also, IR focusing on particular zone classes should improve the quality of retrieval.</Paragraph>
    <Paragraph position="3"> Specifically, the goal of the experiment (a PBM zone) is expected to be used as an index for the organization and retrieval of experimental results.</Paragraph>
  </Section>
class="xml-element"></Paper>