<?xml version="1.0" standalone="yes"?>
<Paper uid="W99-0619">
  <Title>Word Informativeness and Automatic Pitch Accent Modeling</Title>
  <Section position="11" start_page="154" end_page="155" type="concl">
    <SectionTitle>
7 Conclusion
</SectionTitle>
    <Paragraph position="0"> In this paper, we have provided empirical evidence for the usefulness of informativeness for accent assignment. Overall, there is a  positive correlation between indicators of informativeness, such as IC and TF*IDF, and pitch accent. The more informative a word is, the more likely that a pitch accent is assigned to the word. Both of the two measurements of informativeness improve over the baseline performance significantly. We also show that IC is a more powerful measure of informativeness than TF*IDF for pitch accent prediction. Later, when comparing ICempowered POS models with POS models, we found that IC enables additional, statistically significant improvements for pitch accent assignment. This performance also out-performs the TTS pitch accent model significantly. Overall, IC is not only effective, as shown in the results, but also relatively inexpensive to compute for a new domain. Almost all speech synthesis systems, text-to-speech as well as concept-to-speech systems, can employ this feature as long as there is a large corpus. In the future, we plan to explore other information content measurements and incorporate them in a more comprehensive accent model with more discourse and semantic features included.</Paragraph>
  </Section>
class="xml-element"></Paper>