Skip to main navigation Skip to search Skip to main content

Link-sensitive PageRank: An improved ranking algorithm for vertical search engines

  • Ling Zheng*
  • , Ning Zhang
  • , Yang Bo
  • *Corresponding author for this work
  • North China Electric Power University

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

The PageRank algorithm is an important link-based ranking strategy of vertical search engines, but it has a drawback of topic drift. To tackle this problem and yield more accurate search results, we present an improved algorithm to distribute the PageRank value in light of the link sensitive level of the web pages based on keywords set, which we called "Link-Sensitive PageRank". According to the keywords of user's searching, this algorithm, which takes into account the link sensitive level of the web pages' hyperlink to give different importance to different hyperlinks. Experiment results show that the improved PageRank algorithm performs better than the standard PageRank. Futhermore, it can effectively improve the "topic drift" and enhance the accuarcy of information collection. The proposed PageRank algorithm can have a good application in the vertical search engines.

Original languageEnglish
Title of host publication2009 1st International Conference on Information Science and Engineering, ICISE 2009
Pages887-890
Number of pages4
DOIs
StatePublished - 2009
Externally publishedYes
Event1st International Conference on Information Science and Engineering, ICISE2009 - Nanjing, China
Duration: 26 Dec 200928 Dec 2009

Publication series

Name2009 1st International Conference on Information Science and Engineering, ICISE 2009

Conference

Conference1st International Conference on Information Science and Engineering, ICISE2009
Country/TerritoryChina
CityNanjing
Period26/12/0928/12/09

Keywords

  • Keywords
  • Link sensitive
  • PageRank
  • Vertical search engines

Fingerprint

Dive into the research topics of 'Link-sensitive PageRank: An improved ranking algorithm for vertical search engines'. Together they form a unique fingerprint.

Cite this