<?xml version="1.0" standalone="yes"?>
<Paper uid="P04-1030">
  <Title>Head-Driven Parsing for Word Lattices</Title>
  <Section position="7" start_page="0" end_page="0" type="concl">
    <SectionTitle>
6 Conclusions
</SectionTitle>
    <Paragraph position="0"> In this work we present an adaptation of the parsing model of Collins (1999) for application to ASR. The system was evaluated over two sets of data: strings and word lattices. As PARSEVAL measures are not applicable to word lattices, we measured the parsing accuracy using string input. The resulting scores were lower than that original implementation of the model. Despite this, the model was successful as a language model for speech recognition, as measured by WER and ability to extract high-level information. Here, the system performs better than a simple n-gram model trained on the same data, while simultaneously providing syntactic information in the form of parse trees. WER scores are comparable to related works in this area.</Paragraph>
    <Paragraph position="1"> The large size of the parameter set of this parsing model necessarily restricts the size of training data that may be used. In addition, the resource requirements currently present a challenge for scaling up from the relatively sparse word lattices of the NIST HUB-1 corpus (created in a lab setting by professional readers) to lattices created with spontaneous speech in non-ideal conditions. An investigation into the relevant importance of each parameter for the speech recognition task may allow a reduction in the size of the parameter space, with minimal loss of recognition accuracy. A speedup may be achieved, and additional training data could be used. Tuning of parameters using EM has lead to improved WER for other models. We encourage investigation of this technique for lexicalized head-driven lattice parsing.</Paragraph>
  </Section>
class="xml-element"></Paper>