An Algorithm of Topic Distillation Based on Anchor Text

Abstract

For topic distillation and topic exploration, a rich resource is the information from hyper linked Web pages. The anchor text, an objective summary of Web pages, is of strong topic and semantic indication. As to the triangular relationship of the hub Web page, anchor text, and authoritative Web page, a good hub Web page is the one that is directed to a good authoritative Web page through the authoritative anchor text; an authoritative anchor text is the one that connects good hub Web page and good authoritative Web page; and a good authoritative Web page is the one that a good hub Web page directs to through a good authoritative anchor text. This paper is mainly about hypertext-based topic analysis models and community-based topic distillation. Through comparative analysis, it can be concluded that the hypertext-based topic distillation contributes to improving the quality of topic distillation.