Skip to main navigation Skip to search Skip to main content

Exploiting pagerank at different block level

  • Xue-Mei Jiang
  • , Gui-Rong Xue
  • , Wen-Guan Song
  • , Hua-Jun Zeng
  • , Zheng Chen
  • , Ma. Wei-Ying

Research output: Chapters, Conference Papers, Creative and Literary WorksRGC 12 - Chapter in an edited book (Author)peer-review

Abstract

In recent years, information retrieval methods focusing on the link analysis have been developed; The PageRank and HITS arc two typical ones According to the hierarchical organization of Web pages, we could partition the Web graph into blocks at different level, such as page level, directory level, host level and domain level. On the basis of block, we could analyze the different hyperlinks among pages. Several approaches proposed that the intrahyperlink in a host maybe less useful in computing the PageRank. However, there are no reports on how concretely the intra- or inter-hyperlink affects the PageRank. Furthermore, based on different block level, inter-hyperlink and intra-hyperlink can be two relative concepts. Thus which level should be optimal, to distinguish the intra- or inter-hyperlink? And how the ratio set between the intra-hyperlink and inter-hyperlink could ultimately improve performance of the PageRank algorithm? In this paper, we analyze the link distribution at the different block level and evaluate the importance of the intra- and interhyperlink to PageRank on the TREC Web Track data set. Experiment shows that, if we set the block at host level and the ratio of the weight between the intra-hyperlink and inter-hyperlink is 1:4, the retrieval could achieve the best performance. © Springer-Verlag 2004.
Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer Verlag
Pages241-252
Volume3306
ISBN (Print)3540238948, 9783540238942
DOIs
Publication statusPublished - 2004
Externally publishedYes

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3306
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Bibliographical note

Publication details (e.g. title, author(s), publication statuses and dates) are captured on an “AS IS” and “AS AVAILABLE” basis at the time of record harvesting from the data source. Suggestions for further amendments or supplementary information can be sent to [email protected].

Fingerprint

Dive into the research topics of 'Exploiting pagerank at different block level'. Together they form a unique fingerprint.

Cite this