<?xml version="1.0" standalone="yes"?>
<Paper uid="C04-1074">
  <Title>Optimizing Algorithms for Pronoun Resolution</Title>
  <Section position="6" start_page="0" end_page="0" type="concl">
    <SectionTitle>
5 Conclusion
</SectionTitle>
    <Paragraph position="0"> The paper has presented a survey of pronoun resolution factors and algorithms. Two questions were investigated: Which factors should be chosen, and how should they interact? Two types of factors, 'filters' and 'preferences', were discussed in detail.</Paragraph>
    <Paragraph position="1"> In particular, their restrictive potential and effect on success rate were assessed on the evaluation corpus.</Paragraph>
    <Paragraph position="2"> To address the second question, several well-known algorithms were grouped into three classes according to their solution to factor interaction: Serialization, Weighting, and Machine Learning. Six algorithms were evaluated against a common evaluation set so as to facilitate direct comparison. Different algorithms have different strengths, in particular as regards their robustness to parsing errors. Two of the interaction strategies (Serialization and Machine Learning) allow data-driven optimization. Optimal algorithms could be proposed for these strategies.</Paragraph>
  </Section>
class="xml-element"></Paper>