A Novel Approach of Web Search Based on Community Wisdom

Abstract

In this paper, we propose a novel approach for Web search based on the statistical information of local setting data of web browsers in a community. The members of the community share their local setting data of browsers and this enables them to take advantage of the peer community members's opinions in their Web search. Then we develop a new scheme that combines PageRank's link-based ranking scores with our proposed community based popularity scores for web sites. This hybrid scheme provides a rank- ordering method for search query results that integrates the content consumers' opinions with the content producers' opinions in a balanced manner. The users' opinions of web sites provide a solid starting point of trust for combatting web spam and improving the quality of Web search.