<?xml version="1.0" standalone="yes"?>
<Paper uid="A97-1025">
  <Title>Contextual Spelling Correction Using Latent Semantic Analysis</Title>
  <Section position="8" start_page="171" end_page="171" type="concl">
    <SectionTitle>
6 Conclusion
</SectionTitle>
    <Paragraph position="0"> We've shown that LSA can be used to attack the problem of identifying contextual misuses of words, particularly when those words are the same part of speech. It has proven to be an effective alternative to Bayesian classifiers. Confusions sets whose words are different parts of speech are more effectively handled using a method which incorporates the word's part of speech as a feature. We are exploring techniques for introducing part of speech information into the LSA space so that the system can make better predictions for those sets on which it doesn't yet measure up to Tribayes. We've also shown that for the cost of experimentation with different parameter combinations, LSA's performance can be tuned for individual confusion sets.</Paragraph>
    <Paragraph position="1"> While the results of this experiment look very nice, they still don't tell us anything about how useful the technique is when applied to unedited text. The testing procedure assumes that a confusion word must be predicted as if the author of the text hadn't supplied a word or that writers misuse the confusion words nearly 50% of the time. For example, consider the case of the confusion set {principal, principle}.</Paragraph>
    <Paragraph position="2"> The LSA prediction accuracy for this set is 91%.</Paragraph>
    <Paragraph position="3"> However, it might be the case that, in practice, people tend to use the Correct word 95% of the time.</Paragraph>
    <Paragraph position="4"> LSA has thus introduced a 4% error into the writing process. Our continuing work is to explore the error rate that occurs in unedited text as a means of assessing the &amp;quot;true&amp;quot; performance of contextual spelling correction systems.</Paragraph>
  </Section>
class="xml-element"></Paper>