<?xml version="1.0" standalone="yes"?>
<Paper uid="P06-2087">
  <Title>Argumentative Feedback: A Linguistically-motivated Term Expansion for Information Retrieval</Title>
  <Section position="7" start_page="679" end_page="679" type="evalu">
    <SectionTitle>
5 Results and Discussion
</SectionTitle>
    <Paragraph position="0"> All results are computed using the treceval program, using the top 1000 retrieved documents for each evaluation query. We mainly evaluate the impact of varying the feedback category on the retrieval effectiveness, so we separately expand our queries based a single category. Query expansion based on RESULTS or METHODS sentences does not result in any improvement.</Paragraph>
    <Paragraph position="1"> Onthecontrary, expansionbasedonPURPOSE sentences improve the Rocchio baseline by + 23%, which is again significant (p &lt; 0.05). But the main improvement is observed when CONCLUSION sentences are used to generate the expansion, with a remarkable gain of 41% when compared to Rocchio. We also observe in Table 5 that other measures (top precision) and number of relevant retrieved articles do confirm this trend.</Paragraph>
    <Paragraph position="2"> For the PURPOSE category, the optimal k parameter, computed on the test queries was 11. For the CONCLUSION category, the optimal k parameter, computed on the test queries was 10. The difference between the m values between Rocchio feedback and the argumentative feedback, respectively 15 vs. 11 and 10 for Rocchio, PURPOSE, CONCLUSION sentences can  chio and with argumentative feedback applied on PURPOSE and CONCLUSION sentences.</Paragraph>
    <Paragraph position="3"> The number of relevant document for all queries is 1178.</Paragraph>
    <Paragraph position="4"> be explained by the fact that less textual material is available when a particular class of sentences is selected; therefore the number of words that should be added to the original query is more targeted.</Paragraph>
    <Paragraph position="5"> From a more general perspective, the importance of CONCLUSION and PURPOSE sentences is consistent with other studies, which aimed at selecting highly content bearing sentences for information extraction (Ruch et al., 2005b). This result is also consistent with the state-of-the-art in automatic summarization, which tends to prefer sentences appearing at the beginning or at the end of documents to generate summaries.</Paragraph>
  </Section>
class="xml-element"></Paper>