In this paper, based on the classification of search intentions, required information for every intention is further analyzed, and a dynamic model of generalizing Web pages based on the intentions is proposed, which dynamically creates a concept hierarchy concerning snippets, keywords, navigation types, and document formats for searched Web pages, and provides further retrieval navigations for different intentions by generalizing content, format and type of the pages, to return more relevant results to the intentions. Comparing with related work, this paper does not focus on users’ intentions themselves, but on the creation of generalization models based on users’ intentions and implementation for Web pages generalizing. Experimental results show that the generalization model can automatically acquire users’ search intentions under navigation, and provide relevant results and further navigation based on the intentions.