<?xml version="1.0" standalone="yes"?>
<Paper uid="C04-1101">
  <Title>Combining Linguistic Features with Weighted Bayesian Classifier for Temporal Reference Processing</Title>
  <Section position="2" start_page="0" end_page="1" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> Temporal information describes changes and time of the changes. In a language, the time of an event may be specified explicitly, for example &amp;quot;Ta Men Zai 1997 Nian Jie Jue Liao Gai Shi De Jiao Tong Wen Ti (They solved the traffic problem of the city in 1997)&amp;quot;; or it may be related to the time of another event, for example &amp;quot;Xiu Cheng Li Jiao Qiao Yi Hou , Ta Men Jie Jue Liao Gai Shi De Jiao Tong Wen Ti (They solved the traffic problem of the city after the street bridge had been built&amp;quot;. Temporal reference describes how events relate to one another, which is essential to natural language processing (NLP). Its major applications cover syntactic structural disambiguation (Brent, 1990), information extraction and question answering (Li, 2002), language generation and machine translation (Dorr, 2002).</Paragraph>
    <Paragraph position="1"> Many researchers have attempted to characterize the nature of temporal reference in a discourse. Identifying temporal relations  between two events de- null The relations under examined include both intra-sentence and interpends on a combination of information resources.</Paragraph>
    <Paragraph position="2"> This information is provided by explicit tense and aspect markers, implicit event classes or discourse structures. It has been used to explain semantics of temporal expressions (Moens, 1988; Webber, 1988), to constrain possible temporal interpretations (Hitzeman, 1995; Sing, 1997), or to generate appropriate temporally conjoined clauses (Dorr, 2002). The purpose of our work is to develop a computational model, which automatically determines temporal relations in Chinese. While temporal reference interpretation in English has been well studied, Chinese has been rarely discussed. In our study, thirteen related features are identified from linguistic perspective. How to combine these features and how to map their combined effects to the corresponding relations are the critical issues to be addressed in this paper.</Paragraph>
    <Paragraph position="3"> Previous work was limited in that they just constructed constraint or preference rules for some representative examples. These methods are ineffective for computing purpose, especially when a large number of the features are involved and the interaction among them is unclear. Therefore, a machine learning approach is applied and the empirical studies are carried out in our work.</Paragraph>
    <Paragraph position="4"> The rest of this paper is organized as follows. Section 2 introduces temporal relation representations. Section 3 provides linguistic background of temporal reference and investigates linguistic features for determining temporal relations in Chinese. Section 4 explains the methods used to combine linguistic features with Bayesian Classifier. It is followed by a description of the optimization algorithm which is used for estimating feature weights in Section 5. Finally, Section 6 concludes the paper.</Paragraph>
  </Section>
class="xml-element"></Paper>