<?xml version="1.0" standalone="yes"?>
<Paper uid="C02-1097">
  <Title>Word Sense Disambiguation using Static and Dynamic Sense Vectors</Title>
  <Section position="5" start_page="6" end_page="6" type="concl">
    <SectionTitle>
4. Conclusion
</SectionTitle>
    <Paragraph position="0"> This paper reported about word sense disambiguation for English words using static and dynamic sense vectors. Content words noun, verb, and adjective - in the context were selected as contextual words. Local density was used to weight words in the contextual window.</Paragraph>
    <Paragraph position="1"> Then we constructed static sense vectors for each sense. A automatic selective sampling method was used to construct dynamic sense vectors, which had more discriminative power, by reducing the negative effects of noise in the training sense tagged data. The answer was decided by comparing similarity. Our method is simple but effective for WSD.</Paragraph>
    <Paragraph position="2"> Our method leads up to 70~150% precision improvement in the experimentation comparing the system without local density and automatic selective sampling. We showed that our method is simple but effective. Our method was somewhat language independent, because our method used only POS information. Syntactic and semantic features such as dependency relations, approximated word senses of contextual words and so on may be useful to improve the performance of our method.</Paragraph>
  </Section>
class="xml-element"></Paper>