<?xml version="1.0" standalone="yes"?>
<Paper uid="P98-2132">
  <Title>A Multi-Neuro Tagger Using Variable Lengths of Contexts</Title>
  <Section position="7" start_page="805" end_page="805" type="concl">
    <SectionTitle>
6 Conclusion
</SectionTitle>
    <Paragraph position="0"> This paper described a multi-neuro tagger that uses variable lengths of contexts and weighted inputs for part of speech tagging. Computer experiments showed that the multi-neuro tagger has a correct rate of over 94% for tagging ambiguous words when a small Thai corpus with 22,311 ambiguous words is used for training.</Paragraph>
    <Paragraph position="1"> This result is better than any of the results obtained by the single-neuro taggers, which indicates that that the multi-neuro tagger can dynamically find suitable lengths of contexts for tagging. The cost to train a multi-neuro tagger was almost the same as that to train a single-neuro tagger using new learning methods in which the trai~ed results (weights) of the previous taggers are used as initial weights for the latter ones. It was also shown that while the performance of tagging can be improved only slightly, the training time can be greatly reduced by using information gain to weight input elements.</Paragraph>
  </Section>
class="xml-element"></Paper>