<?xml version="1.0" standalone="yes"?>
<Paper uid="W04-1216">
  <Title>Named Entity Recognition in Biomedical Texts using an HMM Model</Title>
  <Section position="7" start_page="121" end_page="121" type="concl">
    <SectionTitle>
6 Conclusion
</SectionTitle>
    <Paragraph position="0"> We trained an HMM model on labelled data to recognize named entities in biomedical texts. Word similarity information was computed from huge unlabeled data. A word similarity-based smoothing method was integrated into the system, and improved the overall performance. We would like to see if it could also be plugged into other existing systems, and hopefully also improve their performance.</Paragraph>
    <Paragraph position="1"> We also argue that the automatically acquired similar words are rich with word features, such as word formation, prefix, suffix, abbreviation, expression variation and clustering information. We will further investigate the usefulness of them in the future.</Paragraph>
  </Section>
class="xml-element"></Paper>