<?xml version="1.0" standalone="yes"?>
<Paper uid="W04-0853">
  <Title>A Gloss-centered Algorithm for Disambiguation</Title>
  <Section position="10" start_page="0" end_page="0" type="evalu">
    <SectionTitle>
5.2 Results for Senseval-3 task
</SectionTitle>
    <Paragraph position="0"> For the Senseval task, we employed hypernym glosses. The remaining parameters and the results are tabulated in table 6.</Paragraph>
    <Paragraph position="1"> We find results quite poor. We performed additional experiments with modified paramater set and find great improvement in numbers. Moreover, we  pick the first WordNet sense in event of lack of any evidence for disambiguation. Hence, in the next reported experiment, the recall values are all same as precision. Based on our experience with the SemCor experiments, we used Hyper-Desc(a9 ) glosses and a context size of 1 sentence. The results are presented in the table 7. The baseline precisions we obtained were by sampling word-senses uniformly at random. The baseline precision was 45.7% for nouns and 25.4% for verbs.</Paragraph>
  </Section>
class="xml-element"></Paper>