<?xml version="1.0" standalone="yes"?>
<Paper uid="W02-0402">
  <Title>Selecting Sentences for Multidocument Summaries using Randomized Local Search</Title>
  <Section position="1" start_page="0" end_page="0" type="abstr">
    <SectionTitle>
Abstract
</SectionTitle>
    <Paragraph position="0"> We present and evaluate a randomized local search procedure for selecting sentences to include in a multidocument summary. The search favors the inclusion of adjacent sentences while penalizing the selection of repetitive material, in order to improve intelligibility without unduly affecting informativeness. Sentence similarity is determined using both surface-oriented measures and semantic groups obtained from merging the output templates of an information extraction subsystem. In a comparative evaluation against two DUC-like baselines and three simpler versions of our system, we found that our randomized local search method provided substantial improvements in both content and intelligibility, while the use of the IE groups also appeared to contribute a small further improvement in content.</Paragraph>
  </Section>
class="xml-element"></Paper>