Using Centrality to Rank Web Snippets

  • Valentin Jijkoun
  • Maarten de Rijke
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5152)

Abstract

We describe our participation in the WebCLEF 2007 task, targeted at snippet retrieval from web data. Our system ranks snippets based on a simple similarity-based centrality, inspired by the web page ranking algorithms. We experimented with retrieval units (sentences and paragraphs) and with the similarity functions used for centrality computations (word overlap and cosine similarity). We found that using paragraphs with the cosine similarity function shows the best performance with precision around 20% and recall around 25% according to human assessments of the first 7,000 bytes of responses for individual topics.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Allan, J., Wade, C., Bolivar, A.: Retrieval and novelty detection at the sentence level. In: SIGIR 2003, pp. 314–321 (2003)Google Scholar
  2. 2.
    Adafre, S.F., Jijkouni, V., de Rijke, M.: Fact discovery in Wikipedia. In: IEEE/WIC/ACM International Conference on Web Intelligence 2007 (2007)Google Scholar
  3. 3.
    Jijkoun, V., de Rijke, M.: Recognizing textual entailment: Is lexical similarity enough? In: Dagan, I., Dalche, F., Quinonero Candela, J., Magnini, B. (eds.) MLCW 2005. LNCS (LNAI), vol. 3944, pp. 449–460. Springer, Heidelberg (2006)Google Scholar
  4. 4.
    Liu, B.: Web Data Mining. Exploring Hyperlinks, Contents and Usage Data. Springer, Heidelberg (2006)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2008

Authors and Affiliations

  • Valentin Jijkoun
    • 1
  • Maarten de Rijke
    • 1
  1. 1.ISLAUniversity of Amsterdam 

Personalised recommendations