|
|
|
                                          
|
Return to Posters
Web link analysis has been proved to provide significant enhancement to the precision of Web search in practice. The PageRank algorithm, which is used in Google Search Engine, plays an important role on improving the quality of its resuls by employing the explicit hyperlink structure among the Web pages. The prestige of Web pages defined by PageRank is purely derived from surfer random walk on the Web graph without textual content content consideration. However, in the practical sense, user surfing behavior is far from random jumping. In this paper, we present a unified model for a more accurate page rank. User's surfing is guided by a probabilistic model that is based on literal matching between connected pages. The result shows that our proposed ranking algorithms do perform better than the original PageRank.
@inproceedings{1009113, author = {Yinghui Xu and Kyoji Umemura}, title = {A unified model of literal mining and link analysis for ranking web resources}, booktitle = {SIGIR '04: Proceedings of the 27th annual international conference on Research and development in information retrieval}, year = {2004}, isbn = {1-58113-881-4}, pages = {546--547}, location = {Sheffield, United Kingdom}, doi = {http://doi.acm.org/10.1145/1008992.1009113}, publisher = {ACM Press}, }  ©2005 Association for Computing Machinery |
