<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-3319">
  <Title>Biomedical Term Recognition With the Perceptron HMM Algorithm</Title>
  <Section position="3" start_page="114" end_page="114" type="evalu">
    <SectionTitle>
2 Results and discussion
</SectionTitle>
    <Paragraph position="0"> We evaluated our system on the JNLPBA Bio-Entity recognition task. The training data set contains 2,000 Medline abstracts labeled with biomedical classes in the IOB style. The IOB annotation method utilizes three types of tags: &lt;B&gt; for the beginning word of a term, &lt;I&gt; for the remaining words of a term, and &lt;O&gt; for non-term words. For the purpose of term classi cation, the IOB tags are augmented with the names of the biomedical classes; for example, &lt;B-protein&gt; indicates the rst word of a protein term. The held-out set was constructed by randomly selecting 10% of the sentences from the available training set. The number of iterations for training was determined by observing the point where the performance on the held-out set starts to level off. The test set is composed of new 404 Medline abstracts.</Paragraph>
    <Paragraph position="1"> Table 1 shows the results of our system on all ve classes. In terms of F-measure, our system achieves  set with respect to each biomedical concept class. the average of 68.6%, which a substantial improvement over the baseline system (based on longest string matching against a lists of terms from training data) with the average of 47.7%, and over the basic HMM system, with the average of 53.9%. In comparison with the results of eight participants at the JNLPBA shared tasks (Kim et al., 2004), our system ranks fourth. The performance gap between our system and the best systems at JNLPBA, which achieved the average up to 72.6%, can be attributed to the use of richer and more complete features such as dictionaries and Gene ontology.</Paragraph>
  </Section>
class="xml-element"></Paper>