<?xml version="1.0" standalone="yes"?>
<Paper uid="P05-1054">
  <Title>A Quantitative Analysis of Lexical Differences Between Genders in Telephone Conversations</Title>
  <Section position="3" start_page="0" end_page="435" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> Linguistic and prosodic differences between genders in American English have been studied for decades. The interest in analyzing the gender linguistic differences is two-fold. From the scientific perspective, it will increase our understanding of language production. From the engineering perspective, it can help improve the performance of a number of natural language processing tasks, such as text classification, machine translation or automatic speech recognition by training better language models. Traditionally, these differences have been investigated in the fields of sociolinguistics and psycholinguistics, see for example (Coates, 1997), (Eckert and McConnell-Ginet, 2003) or http://www.ling.lancs.ac.uk/groups/gal/genre.htm for a comprehensive bibliography on language and gender. Sociolinguists have approached the issue from a mostly non-computational perspective using relatively small and very focused data collections.</Paragraph>
    <Paragraph position="1"> Recently, the work of (Koppel et al., 2002) has used computational methods to characterize the differences between genders in written text, such as literary books. A number of monologues have been analyzed in (Singh, 2001) in terms of lexical richness using multivariate analysis techniques.</Paragraph>
    <Paragraph position="2"> The question of gender linguistic differences shares a number of issues with stylometry and author/speaker attribution research (Stamatatos et al., 2000), (Doddington, 2001), but novel issues emerge with analysis of conversational speech, such as studying the interaction of genders.</Paragraph>
    <Paragraph position="3"> In this work, we focus on lexical differences between genders on telephone conversations and use machine learning techniques applied on text categorization and feature selection to characterize these differences. Therefore our conclusions are entirely data-driven. We use a very large corpus created for automatic speech recognition - the Fisher corpus described in (Cieri et al., 2004). The Fisher corpus is annotated with the gender of each speaker making it an ideal resource to study not only the characteristics of individual genders but also of gender pairs in spontaneous, conversational speech. The size and  scope of the Fisher corpus is such that robust results can be derived for American English. The computational methods we apply can assist us in answering questions, such as &amp;quot;To which degree are gender-discriminative words content-bearing words?&amp;quot; or &amp;quot;Which words are most characteristic for males in general or males talking to females?&amp;quot;.</Paragraph>
    <Paragraph position="4"> In section 2, we describe the corpus we have based our analysis on. In section 3, the machine learning tools are explained, while the experimental results are described in section 4 with a specific research question for each subsection. We conclude in section 5 with a summary and future directions.</Paragraph>
  </Section>
class="xml-element"></Paper>