<?xml version="1.0" standalone="yes"?>
<Paper uid="I05-2025">
  <Title>Investigating the features that affect cue usage of non-native speakers of English</Title>
  <Section position="2" start_page="0" end_page="144" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> As an international language, English has become more and more important for non-native speakers. However, almost all English documents are written for the native speakers. To some degree, some documents can not be understood quite well by non-native speakers. This paper concentrates on exploring the differences in cue usage at discourse level between native and non-native speakers. The aim is to find the decision-making mechanisms of text generation for users at different reading levels.</Paragraph>
    <Paragraph position="1"> While investigating texts written for non-native speakers, we found that cue phrase because sometimes occurs in the first span of a discourse relation. This is different from the conclusion mentioned in (Quirk and Greenbaum and Leech and Svartvik, 1972), that is, (for native speakers) because typically occurs in the second span. This problem could be considered from the viewpoint of text generation as well. The following three texts may have the same abstract text structure, though the differences among them are apparent.</Paragraph>
    <Paragraph position="2"> E.g., cue placement is different. In text (1), cue phrase because occurs at first span of discourse relation &amp;quot;explanation&amp;quot;, while in (2) and (3), because occurs in the second span.</Paragraph>
    <Paragraph position="3">  Example 1.1: 1. Global warming will be a major threat to the  whole world over the next century. But because it will take many years for our actions to produce a significant effect, the problem needs attention now.</Paragraph>
    <Paragraph position="4"> 2. Global warming will be a major threat to the whole world over the next century, but the problem needs attention now, because it will take many years for our actions to produce a significant effect.</Paragraph>
    <Paragraph position="5"> 3. Global warming will be a major threat to the whole world over the next century. But the problem needs attention now, because it will take many years for our actions to produce a significant effect.</Paragraph>
    <Paragraph position="6"> This paper reports the results of the research on the different placement (where to place a cue) of because between native and non-native speakers through analyzing two annotated corpora. At the same time, we study the features that affect placement of because for non-native speakers. The rest of the paper is arranged as follows. Section 2 describes related work. Section 3 demonstrates how to create two corpora (SUB-BNC and CNNSE).</Paragraph>
    <Paragraph position="7">  Section 4 shows the method of annotating corpora. Section 5 demonstrates the difference in usage of because between two corpora. In section 6, a machine learning program - C4.5 is introduced. Section 7 shows the experimental results. Section</Paragraph>
  </Section>
class="xml-element"></Paper>