Abstract: This paper improves PageRank algorithm is based on analyzing the phenomenon of stressing on old pages, drifting theme, spliting page weight and neglecting user browsing interests. It considers the important factors of webpage modification data, webpage text information, website technoroti authority, user interestingness. Experimental result shows that improved algorithm can improve accuracy for webpage order and user satisfaction with search results.

Key words: PageRank algorithm, search engine, text data mining, PR value

摘要： 在分析PageRank算法存在偏重旧网页、主题漂移、网页权值均分、忽视用户浏览兴趣现象的基础上，对其进行改进，考虑网页修改日期、网页文本信息、网站权威度、用户兴趣度等重要因素，重新计算网页PR值。实验结果表明，改进算法可提高搜索引擎对网页排序的准确度，以及用户对检索结果的满意度。

关键词: PageRank算法, 搜索引擎, 文本数据挖掘, PR值

CLC Number: 

• TP393