<?xml version="1.0" standalone="yes"?>
<Paper uid="P06-2013">
  <Title>Sydney, July 2006. c(c)2006 Association for Computational Linguistics An Empirical Study of Chinese Chunking</Title>
  <Section position="10" start_page="102" end_page="102" type="concl">
    <SectionTitle>
7 Conclusions
</SectionTitle>
    <Paragraph position="0"> In this paper, we conducted an empirical study of Chinese chunking. We compared the performance of four models, SVMs, CRFs, MBL, and TBL.</Paragraph>
    <Paragraph position="1"> We also investigated the effects of using different sizes of training data. In order to provide higher accuracy, we proposed two new voting methods according to the characteristics of the chunking task. We proposed the Tag-Extension approach to resolve the special problems of Chinese chunking by extending the chunk tags.</Paragraph>
    <Paragraph position="2"> The experimental results showed that the SVMs model was superior to the other three models.</Paragraph>
    <Paragraph position="3"> We also found that part-of-speech tags played an important role in Chinese chunking because the gap of the performance between WORD+POS and POS was very small.</Paragraph>
    <Paragraph position="4"> We found that the proposed voting approaches can provide higher accuracy than any single system can. In particular, the Phrase-based Voting approach is more suitable for chunking task than the other two voting approaches. Our experimental results also indicated that the Tag-Extension approach can improve the performance significantly.</Paragraph>
  </Section>
class="xml-element"></Paper>