<?xml version="1.0" standalone="yes"?>
<Paper uid="P98-2237">
  <Title>Using Chunk Based Partial Parsing of Spontaneous Speech in Unrestricted Domains for Reducing Word Error Rate in Speech Recognition</Title>
  <Section position="4" start_page="0" end_page="1453" type="intro">
    <SectionTitle>
2 Chunk Parsing
</SectionTitle>
    <Paragraph position="0"> There have been recent developments which encourage the investigation of the possibility of parsing speech in unrestricted domains. It was demonstrated that parsing natural language 2 can be han-WER ----- 100.0. substitutiona-~d~leticms-~insertions correctt ~ubstitutiollsJrdC/|C/tion$  dled by very simple, even finite-state approaches if one adheres to the principle of &amp;quot;chunking&amp;quot; the input into small and hence easily manageable constituents (Abney, 1996; Light, 1996).</Paragraph>
    <Paragraph position="1"> We use the notion of a chunk similar to (Abney, 1996), namely a contiguous, non-recursive phrase.</Paragraph>
    <Paragraph position="2"> Chunk phrases mostly correspond to traditional notions of syntactic constituents, such as NPs or PPs, but there are exceptions, e.g. VCs (&amp;quot;verb complex phrases&amp;quot;), which are not used in most traditional linguistic paradigms. 3 Unlike in (Abney, 1996), our goal was not to build a multi-stage, cascaded system to result in full sentence parses, but to confine ourselves to parsing of &amp;quot;basic chunks&amp;quot;.</Paragraph>
    <Paragraph position="3"> A strong rationale for following this simple approach is the nature of the ill-formed input due to (i) spontaneous speech dysfluencies, and (ii) errors in the hypotheses of the speech recognizer.</Paragraph>
    <Paragraph position="4"> To get an intuitive feel about the output of the chunk parser, we present a short example here: 4 \[conj BUT\] \[np HE\] \[vc DOESN'T REALLY LIKE\] \[np HIS HISTORY TEACHER\] \[advp VERY MUCH\]</Paragraph>
  </Section>
class="xml-element"></Paper>