<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-1317">
  <Title>Classification of Discourse Coherence Relations: An Exploratory Study using Multiple Knowledge Sources</Title>
  <Section position="4" start_page="0" end_page="117" type="intro">
    <SectionTitle>
2 Previous Work
</SectionTitle>
    <Paragraph position="0"> In the past few years, the tasks of discourse segmentation and parsing have been tackled from different perspectives and within different frameworks. Within Rhetorical Structure Theory (RST), Soricut and Marcu (2003) have developed two  probabilistica5 models for identifying clausal elementary discourse units and generating discourse trees at the sentence level. These are built using lexical and syntactic information obtained from mapping the discourse-annotated sentences in the RST Corpus (Carlson et al., 2003) to their corresponding syntactic trees in the Penn Treebank. Within SDRT, Baldridge and Lascarides (2005b) also take a data-driven approach to the tasks of segmentation and identification of discourse relations. They create a probabilistic discourse parser based on dialogues from the Redwoods Treebank, annotated with SDRT rhetorical relations (Baldridge and Lascarides, 2005a). The parser is grounded on headed tree representations and dialogue-based features, such as turn-taking and domain specific goals.</Paragraph>
    <Paragraph position="1"> In the Penn Discourse TreeBank (PDTB) (Webber et al., 2005), the identification of discourse structure is approached independently of any linguistic theory by using discourse connectives rather than abstract rhetorical relations. PDTB assumes that connectives are binary discourse-level predicates conveying a semantic relationship between two abstract object-denoting arguments.</Paragraph>
    <Paragraph position="2"> The set of semantic relationships can be established at different levels of granularity, depending on the application. Miltsakaki, et al. (2005) propose a first step at disambiguating the sense of a small subset of connectives (since, while, and when) at the paragraph level. They aim at distinguishing between the temporal, causal, and contrastive use of the connective, by means of syntactic features derived from the Penn Treebank and a MaxEnt model.</Paragraph>
  </Section>
class="xml-element"></Paper>