<?xml version="1.0" standalone="yes"?>
<Paper uid="A00-1043">
  <Title>D B E G /\ /',, A C F H ? A D t B E G C F H Reduced: A B D G H Figure 3: Reduced form by a human D B E G A C F H Reduced: B C D Input: A B C D E F G H Figure 4: Reduced form by the program</Title>
  <Section position="6" start_page="314" end_page="314" type="concl">
    <SectionTitle>
5 Conclusions and future work
</SectionTitle>
    <Paragraph position="0"> We present a novel sentence reduction system which removes extraneous phrases from sentences that are extracted from an article in text summarization. The deleted phrases can be prepositional phrases, clauses, to-infinitives, or gerunds, and multiple phrases can be removed form a single sentence. The focus of this work is on determining, for a sentence in a particular context, which phrases in the sentence are less important and can be removed. Our system makes intelligent reduction decisions based on multiple sources of knowledge, including syntactic knowledge, context, and probabilities computed from corpus analysis. We also created a corpus consisting of 500 sentences and their reduced forms produced by human professionals, and used this corpus for training and testing the system. The evaluation shows that 81.3% of reduction decisions made by the system agreed with those of humans.</Paragraph>
    <Paragraph position="1"> In the future, we would like to integrate our sentence reduction system with extraction-based summarization systems other than the one we have developed, improve the performance of the system further by introducing other sources of knowledge necessary for reduction, and explore other interesting applications of the reduction system.</Paragraph>
  </Section>
class="xml-element"></Paper>