<?xml version="1.0" standalone="yes"?>
<Paper uid="H92-1021">
  <Title>IMPROVEMENTS IN STOCHASTIC LANGUAGE MODELING</Title>
  <Section position="3" start_page="0" end_page="0" type="intro">
    <SectionTitle>
1. INTRODUCTION
</SectionTitle>
    <Paragraph position="0"> Linguistic constraints are an important factor in human comprehension of speech. Their effect on automatic speech recognition is similar, in that they provide both a pruning method and a means of ordering likely candidates. As vocabularies for speech recognition systems increase in size, more accurate modeling of linguistic constraints becomes essential.</Paragraph>
    <Paragraph position="1"> Two fundamental issues in language modeling are smoothing and adaptation. Smoothing allows a model to assign reasonable probabilities to events that have never been observed before. Adaptation takes advantage of recently gained knowledge -- the text seen so far -- to adjust the model's expectations.</Paragraph>
    <Paragraph position="2"> In what follows, we discuss two attempts at improving our current stochastic language modeling techniques. In the first, we try to improve smoothing by correcting a deficiency in a successful and well known smoothing method, the backoff model. In the second, we propose a novel kind of adaptation, one that is based on correlation among word sequences occurring in the same document.</Paragraph>
  </Section>
class="xml-element"></Paper>