<?xml version="1.0" standalone="yes"?>
<Paper uid="E99-1031">
  <Title>A Flexible Architecture for Reference Resolution</Title>
  <Section position="2" start_page="0" end_page="229" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> When building natural language understanding systems, choosing the best technique for anaphora resolution is a challenging task. The system builder must decide whether to adopt an existing technique or design a new approach.</Paragraph>
    <Paragraph position="1"> A huge variety of techniques are described in the literature, many of them achieving high success rates on their own evaluation texts (cf.</Paragraph>
    <Paragraph position="2"> Hobbs 1986; Strube 1998; Mitkov 1998). Each technique makes different assumptions about the data available to reference resolution, for example, some assume perfect parses, others assume only POS-tagged input, some assume semantic information is available, etc. The chances are high that no published technique will exactly match the data available to a particular system's reference resolution component, so it may The authors thank James Allen for help on this project, as well as the anonymous reviewers for helpful comments on the paper. This material is based on work supported by USAF/Rome Labs contract F30602-95-1-0025, ONR grant N00014-95-1 - 1088, and Columbia Univ. grant OPG: 1307.  not be apparent which method will work best.</Paragraph>
    <Paragraph position="3"> Choosing a technique is especially problematic for designers of dialogue systems trying to predict how anaphora resolution techniques developed for written monologue will perform when adapted for spoken dialogue. In an ideal world, the system designer would implement and compare many techniques on the input data available in his system. As a good software engineer, he would also ensure that any pronoun resolution code he implements can be ported to future applications or different language domains without modification.</Paragraph>
    <Paragraph position="4"> The architecture described in this paper was designed to provide just that functionality.</Paragraph>
    <Paragraph position="5"> Anaphora resolution code developed within the architecture is encapsulated to ensure portability across parsers, language genres and domains. Using these architectural guidelines, a testbed system for comparing pronoun resolution techniques has been developed at the University of Rochester. The testbed provides a highly configurable environment which uses the same pronoun resolution code regardless of the parser front-end and language type under analysis. It can be used, inter alia, to compare anaphora resolution techniques for a given application, to compare new techniques to published baselines, or to compare a particular technique's performance across language types.</Paragraph>
  </Section>
class="xml-element"></Paper>