<?xml version="1.0" standalone="yes"?>
<Paper uid="P05-1017">
  <Title>Extracting Semantic Orientations of Words using Spin Model</Title>
  <Section position="2" start_page="0" end_page="0" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> Identification of emotions (including opinions and attitudes) in text is an important task which has a variety of possible applications. For example, we can efficiently collect opinions on a new product from the internet, if opinions in bulletin boards are automatically identified. We will also be able to grasp people's attitudes in questionnaire, without actually reading all the responds.</Paragraph>
    <Paragraph position="1"> An important resource in realizing such identification tasks is a list of words with semantic orientation: positive or negative (desirable or undesirable). Frequent appearance of positive words in a document implies that the writer of the document would have a positive attitude on the topic. The goal of this paper is to propose a method for automatically creating such a word list from glosses (i.e., definition or explanation sentences ) in a dictionary, as well as from a thesaurus and a corpus. For this purpose, we use spin model, which is a model for a set of electrons with spins. Just as each electron has a direction of spin (up or down), each word has a semantic orientation (positive or negative). We therefore regard words as a set of electrons and apply the mean field approximation to compute the average orientation of each word. We also propose a criterion for parameter selection on the basis of magnetization, a notion in statistical physics. Magnetization indicates the global tendency of polarization.</Paragraph>
    <Paragraph position="2"> We empirically show that the proposed method works well even with a small number of seed words.</Paragraph>
  </Section>
class="xml-element"></Paper>