<?xml version="1.0" standalone="yes"?>
<Paper uid="C02-1081">
  <Title>Data-driven Classification of Linguistic Styles in Spoken Dialogues</Title>
  <Section position="7" start_page="0" end_page="0" type="evalu">
    <SectionTitle>
5.2 Results
</SectionTitle>
    <Paragraph position="0"> When interpreting the results one has to keep a few things in mind: a4 The assumption that the correct style class of a speaker is known in advance is not likely to be true in real systems. A few turns have to be analyzed in order to perform a reasonable classification.</Paragraph>
    <Paragraph position="1"> a4 The parameters used for classification (distributions of part-of-speech items, length parameters etc.) are only very loosely related to the probability of word sequences.</Paragraph>
    <Paragraph position="2"> a4 The classes are not optimized to yield maximal gain in perplexity.</Paragraph>
    <Paragraph position="3"> a4 The language models are rather simple.</Paragraph>
    <Paragraph position="4"> The global results are displayed in Figure 8. The general model (with 4 times the training material than the special models) gives better results, except for the TABA corpus which is probably sufficiently constrained and simply structured to make up for the decrease in training material. This is in line with results described in Klarner (1997). The interpolated model has a significantly lower perplexity than the general model alone, but the gain is  runs (n=324) between style-specific and general language models (top) and between interpolated and general language models (bottom). Significance was calculated by the paired t-test (one-sided, a6a8a7a10a9a12a11a9a14a13 , a15a17a16a19a18a21a20a22a13 ). so small that it is unlikely to improve recognition results. With all the caveats listed above one can conclude that determination of linguistic style in the way described in this document does not dramatically improve recognition results.</Paragraph>
  </Section>
class="xml-element"></Paper>