基于PageRank与HITS的改进算法的网页排名优化

针对传统网页排序算法PageRank和HITS中存在的主题漂移、检索效率低等不足,本文提出了一种改进算法PHIA(PageRank and HITS Improved Algorithm)。该算法继承了HITS算法获取根集和基本集的方法,并且使用根集中所有网页的PageRank值作为Hub和Authority初始迭代值,最后根据马尔可夫链求随机矩阵的特征向量的方式来获取网页排名的静态分布。基于随机关键词的检索结果可知,相比于传统的PageRank和HITS算法,改进PHIA算法具有更快的收敛速度,并且在一定程度上提高了网页排序的准确度。

An improved algorithm for page rank optimization based on PageRank and HITS algorithms

Aiming at overcoming the disadvantages such as topic drift and low retrieval efficiency in the traditional webpage ranking algorithms PageRank and HITS, an improved algorithm named PHIA (PageRank and HITS Improved Algorithm) was proposed. Firstly, the algorithm inherits the way of HITS algorithm to obtain the root set and the basic set, then employs the PageRank value of all web pages in the root set as the initial iteration value of Hub and Authority, and finally, the page ranking status is obtained by searching the eigenvectors of random matrix based on the Markov chain. The calculation results based on random keyword retrieval show that compared with the traditional PageRank and HITS algorithms, the improved PHIA algorith not only has a faster convergence rate but also improves the accuracy of page ranking to some extent.