【摘要】 首先分析CiteSeer引文网络的结构特征,然后深入讨论PageRank算法的本质,并在此基础上给出一种科学文献个性化排序算法;最后将该算法应用于CiteSeer引文网络,并对实验结果进行分析。个性化向量的计算是实现个性化排序的关键,本文给出等概率、伪被引次数和带时间约束的伪被引次数3种计算方法。更多还原

【Abstract】 Inspired by the PageRank algorithm,a personalized ranking algorithm for scientific publications is discussed in this paper.Three methods for computing the personalization vector are presented.These methods are based on not only user ratings but also citation counts and publication dates.The personalized algorithm is then applied to the citation graph of CiteSeer digital library and the experiment result is analyzed.In addition,the structure of the citation graph is fully explored.更多还原