<?xml version="1.0" standalone="yes"?>
<Paper uid="P06-1032">
  <Title>Correcting ESL Errors Using Phrasal SMT Techniques</Title>
  <Section position="3" start_page="0" end_page="249" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> Every day, in schools, universities and businesses around the world, in email and on blogs and websites, people create texts in languages that are not their own, most notably English. Yet, for writers of English as a Second Language (ESL), useful editorial assistance geared to their needs is surprisingly hard to come by. Grammar checkers such as that provided in Microsoft Word have been designed primarily with native speakers in mind. Moreover, despite growing demand for ESL proofing tools, there has been remarkably little progress in this area over the last decade. Research into computer feedback for ESL writers remains largely focused on small-scale pedagogical systems implemented within the framework of CALL (Computer Aided Language Learning) (Reuer 2003; Vanderventer Faltin, 2003), while commercial ESL grammar checkers remain brittle and difficult to customize to meet the needs of ESL writers of different first-language (L1) backgrounds and skill levels.</Paragraph>
    <Paragraph position="1"> Some researchers have begun to apply statistical techniques to identify learner errors in the context of essay evaluation (Chodorow &amp; Leacock, 2000; Lonsdale &amp; Strong-Krause, 2003), to detect non-native text (Tomokiyo &amp; Jones, 2001), and to support lexical selection by ESL learners through first-language translation (Liu et al., 2000). However, none of this work appears to directly address the more general problem of how to robustly provide feedback to ESL writers--and for that matter non-native writers in any second language--in a way that is easily tailored to different L1 backgrounds and second-language (L2) skill levels.</Paragraph>
    <Paragraph position="2"> In this paper, we show that a noisy channel model instantiated within the paradigm of Statistical Machine Translation (SMT) (Brown et al., 1993) can successfully provide editorial assistance for non-native writers. In particular, the SMT approach provides a natural mechanism for suggesting a correction, rather than simply stranding the user with a flag indicating that the text contains an error. Section 2 further motivates the approach and briefly describes our SMT system. Section 3 discusses the data used in our experiment, which is aimed at repairing a common type of ESL error that is not well-handled by current grammar checking technology: mass/count noun confusions. Section 4 presents experimental results, along with an analysis of errors produced by the system. Finally we present discussion and some future directions for investigation.</Paragraph>
  </Section>
class="xml-element"></Paper>