Integration of information filtering and data mining process for web information retrieval
Xujuan Zhou, Yuefeng Li, Peter Bruza, Yue Xu · QUT ePrints (Queensland University of Technology) · 2007
Abstract This paper examines a new approach to Web information retrieval, and proposes a new two stage scheme. The aim of the first stage is to quickly filter irrelevant information based on the user profiles. The proposed user profiles learning algorithm are very efficient and effective within a relevance feedback framework. The aim of the second stage is to apply data mining techniques to rationalize the data relevance on the reduced data set. Our experiments on RCV1 (Reuters Corpus Volume 1) data collection which is used by TREC in 2002 for filtering track show that more effective and efficient access Web information has been achieved by combining the strength of information filtering and data mining method.