<?xml version="1.0" standalone="yes"?>
<Paper uid="A00-2024">
  <Title>Cut and Paste Based Text Summarization</Title>
  <Section position="9" start_page="184" end_page="184" type="concl">
    <SectionTitle>
7 Conclusions and future work
</SectionTitle>
    <Paragraph position="0"> This paper presents a novel architecture for text summarization using cut and paste techniques observed in human-written abstracts. In order to automatically analyze a large quantity of human-written abstracts, we developed a decomposition program.</Paragraph>
    <Paragraph position="1"> The automatic decomposition allows us to build large corpora for studying sentence reduction and sentence combination, which are two effective operations in cut and paste. We developed a sentence reduction module that makes reduction decisions using multiple sources of knowledge. We also investigated possible sentence combination operations and implemented the combination module. A sentence extraction module was developed and used as the front end of the summarization system.</Paragraph>
    <Paragraph position="2"> We are preparing the task-based evaluation of the overall system. We also plan to evaluate the portability of the system by testing it on another corpus. We will also extend the system to query-based summarization and investigate whether the system can be modified for multiple document summarization.</Paragraph>
  </Section>
class="xml-element"></Paper>