<?xml version="1.0" standalone="yes"?>
<Paper uid="C04-1145">
  <Title>Morpheme-based Derivation of Bipolar Semantic Orientation of Chinese Words</Title>
  <Section position="3" start_page="0" end_page="0" type="relat">
    <SectionTitle>
2. Related Work and Applications
</SectionTitle>
    <Paragraph position="0"> Hatzivassiloglou and Mckeown (1997) presented a method for automatically assigning a + or orientation label to adjectives known to have some SO by the linguistic constraints on the use of adjectives in conjunctions. For example, 'and' links adjectives that have the same SO, while 'but' links adjectives that have opposite SO. They devised an algorithm based on such constraints to evaluate 1,336 manually-labeled adjectives (657 positive and 679 negative) with 97% accuracy in a corpus of 21 million words.</Paragraph>
    <Paragraph position="1"> Turney (2003) introduced a method for automatically inferring the direction and intensity of the SO of a word from its statistical association with a set of positive and negative paradigm words, i.e., strongly-polarized words. The algorithm was evaluated on 3,596 words (1,614 positive and 1,982 negative) including adjectives, adverbs, nouns, and verbs. An accuracy of 82.8% was attained in a corpus of hundred billion words.</Paragraph>
    <Paragraph position="2"> SO can be used to classify reviews (e.g., movie reviews) as positive or negative (Turney, 2002), and applied to subjectivity analysis such as recognizing hostile messages, classifying emails, mining reviews (Wiebe et al., 2001). The first step of those applications is to recognize that the text is subjective and then the second step, naturally, is to determine the SO of the subjective text. Also, it can be used to summarize argumentative articles like editorials of news media. A summarization system would benefit from distinguishing sentences intended to present factual materials from those intended to present opinions, since many summaries are meant to include only facts.</Paragraph>
  </Section>
class="xml-element"></Paper>