<?xml version="1.0" standalone="yes"?>
<Paper uid="P06-1005">
  <Title>Bootstrapping Path-Based Pronoun Resolution</Title>
  <Section position="3" start_page="0" end_page="0" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> Pronoun resolution is a dif cult but vital part of the overall coreference resolution task. In each of the following sentences, a pronoun resolution system  must determine what the pronoun his refers to: (1) John needs his friend.</Paragraph>
    <Paragraph position="1"> (2) John needs his support.</Paragraph>
    <Paragraph position="2">  In (1), John and his corefer. In (2), his refers to some other, perhaps previously evoked entity. Traditional pronoun resolution systems are not designed to distinguish between these cases. They lack the speci c world knowledge required in the second instance the knowledge that a person does not usually explicitly need his own support. We collect statistical path-coreference information from a large, automatically-parsed corpus to address this limitation. A dependency path is dened as the sequence of dependency links between two potentially coreferent entities in a parse tree. A path does not include the terminal entities; for example, John needs his support and He needs their support have the same syntactic path. Our algorithm determines that the dependency path linking the Noun and pronoun is very likely to connect coreferent entities for the path Noun needs pronoun's friend, while it is rarely coreferent for the path Noun needs pronoun's support.</Paragraph>
    <Paragraph position="3"> This likelihood can be learned by simply counting how often we see a given path in text with an initial Noun and a nal pronoun that are from the same/different gender/number classes. Cases such as John needs her support or They need his support are much more frequent in text than cases where the subject noun and pronoun terminals agree in gender/number. When there is agreement, the terminal nouns are likely to be coreferent. When they disagree, they refer to different entities. After a suf cient number of occurrences of agreement or disagreement, there is a strong statistical indication of whether the path is coreferent (terminal nouns tend to refer to the same entity) or non-coreferent (nouns refer to different entities). We show that including path coreference information enables signi cant performance gains on three third-person pronoun resolution experiments. We also show that coreferent paths can provide the seed information for bootstrapping other, even more important information, such as the gender/number of noun phrases.</Paragraph>
  </Section>
class="xml-element"></Paper>