摘要
通过对网页用户角色的分析发现,传统的基于PageRank算法的搜索引擎结果排序欠佳,是因为其没有兼顾所有角色对网页重要性的评价。为此,提出一种结合了所有角色评价的综合网页排序算法——ComPageRank(CPR)算法和一种基于点击量分析的Click-throughRank(CTR)算法。实验结果表明,相比PageRank为代表的网页排序算法,CPR算法更全面、合理。
According to the analysis of the user roles,this paper uncovers one reason of the disadvantage of PageRank–based searching engines,that is,lack of the evaluation of the pages' importance from the point of view of all the roles.A comprehensive page ranking algorithm——ComPageRank(CPR) is proposed.The algorithm takes account the evaluation of all the roles.In addition,a Click-ThroughRank(CTR) algorithm is developed to analyze the click-through.Experimental results show that CPR is more comprehensive and reasonable compared with the PageRank-based algorithms.
出处
《计算机工程》
CAS
CSCD
北大核心
2011年第7期53-55,共3页
Computer Engineering
基金
上海市自然科学基金资助项目(10ZR1421100)
上海市教委创新基金资助项目(08YZ98)