The Static Ranking Algorithm Of Web Pages Based On The Importance Propagation Model

Its data have characteristics of no structure, great quantity, containing hyperlink and updating quickly, moreover the quality good and evil intermingled. Therefore, Web users rely on the search engine more and more when retrieving information. Thereby the results, returned by the search engine, needs to be correct, correlative and ranked. The traditional contents information retrieval technology does not adapt in the huge Web information source. The link analysis technology based on the Web structure excavation brings the Web information retrieval brand-new thinking. The link analysis technology has been used in the search engine not only for sorting the search results (dynamic ranking), but also for sorting the entire Web page (static ranking). Static ranking has important significance in improving the efficiency of search engine, selecting prior crawling order and other aspects.In this paper, based on elaborating the theories of Web data mining, and detailed analyzed and compared the existing link analysis algorithm, thoroughly analyzed the good and bad points of the PageRank algorithm, a static ranking algorithm, and its improved algorithms, it proposed a static ranking algorithm framework of Web pages based on importance propagation model. The PageRank algorithm and its improved algorithms just only consider the important propagation in the pages have direct links relations. The proposed static ranking algorithm framework takes the direct and indirect effects between Web pages into account, so it accurately represents the recommendation of the hyper links. When the important propagation distance is 1, this algorithm framework degenerates into PageRank algorithm.According to the social property of the Web, two instances of the algorithm framework had been given in this paper by using the theory of the attraction among the residential spots and the theory of the distribution of inhabited area. At the same time, it concrete analysis the effective distance of the page importance propagation. Experimental results show that static ranking algorithm framework of Web pages based on importance propagation model not only effectively improves the search precision, but also greatly speeds up the page ranking.