<?xml version="1.0" standalone="yes"?>
<Paper uid="P06-1049">
  <Title>Sydney, July 2006. c(c)2006 Association for Computational Linguistics A Bottom-up Approach to Sentence Ordering for Multi-document Summarization</Title>
  <Section position="4" start_page="0" end_page="385" type="relat">
    <SectionTitle>
2 Related Work
</SectionTitle>
    <Paragraph position="0"> Existing methods for sentence ordering are divided into two approaches: making use of chronological information (McKeown et al., 1999; Lin  and Hovy, 2001; Barzilay et al., 2002; Okazaki et al., 2004); and learning the natural order of sentences from large corpora not necessarily based on chronological information (Lapata, 2003; Barzilay and Lee, 2004). A newspaper usually disseminates descriptions of novel events that have occurred since the last publication. For this reason, ordering sentences according to their publication date is an effective heuristic for multidocument summarization (Lin and Hovy, 2001; McKeown et al., 1999). Barzilay et al. (2002) have proposed an improved version of chronological ordering by first grouping sentences into sub-topics discussed in the source documents and then arranging the sentences in each group chronologically.</Paragraph>
    <Paragraph position="1"> Okazaki et al. (2004) have proposed an algorithm to improve chronological ordering by resolving the presuppositional information of extracted sentences. They assume that each sentence in newspaper articles is written on the basis that presuppositional information should be transferred to the reader before the sentence is interpreted. The proposed algorithm first arranges sentences in a chronological order and then estimates the presuppositional information for each sentence by using the content of the sentences placed before each sentence in its original article. The evaluation results show that the proposed algorithm improves the chronological ordering significantly.</Paragraph>
    <Paragraph position="2"> Lapata (2003) has suggested a probabilistic model of text structuring and its application to the sentence ordering. Her method calculates the transition probability from one sentence to the next from a corpus based on the Cartesian product between two sentences defined using the following features: verbs (precedent relationships of verbs in the corpus); nouns (entity-based coherence by keeping track of the nouns); and dependencies (structure of sentences). Although she has not compared her method with chronological ordering, it could be applied to generic domains, not relying on the chronological clue provided by newspaper articles.</Paragraph>
    <Paragraph position="3"> Barzilay and Lee (2004) have proposed content models to deal with topic transition in domain specific text. The content models are formalized by Hidden Markov Models (HMMs) in which the hidden state corresponds to a topic in the domain of interest (eg, earthquake magnitude or previous earthquake occurrences), and the state transitions capture possible information-presentation orderings. The evaluation results showed that their method outperformed Lapata's approach by a wide margin. They did not compare their method with chronological ordering as an application of multi-document summarization.</Paragraph>
    <Paragraph position="4"> As described above, several good strategies/heuristics to deal with the sentence ordering problem have been proposed. In order to integrate multiple strategies/heuristics, we have formalized them in a machine learning framework and have considered an algorithm to arrange sentences using the integrated strategy.</Paragraph>
  </Section>
class="xml-element"></Paper>