【摘要】 经典的PageRank算法主要根据网页间链接结构进行分析排序,容易产生主题漂移、偏重旧网页、平均分配网页权值、忽视用户兴趣等问题。为了改进上述缺陷,提出一种基于两级修正的页面排序改进算法。综合利用网页相似度、点击量趋势、站内外链接、时间因子对网页的排序值进行一级修正;在此基础上结合用户兴趣对查询结果的排序值进行二级修正,使网页重要性权值计算更为合理;组织毕业生、在校生、教师3组人群进行实验调查,实验结果表明,与传统Page-Rank算法相比较,改进后算法的排序结果查准率更高,更满足用户的需求。更多还原

【Abstract】 Classic PageRank algorithm analyzed the sort of pages mainly based on link structure of web pages,so this algorithm easily lead to topic-drift,emphasis on old pages,evenly distributed web weight,ignorunce of the user interest and other issues. To overcome these defects,an improved ranking algorithm based on two-stage correction was proposed. Firstly,using the web similarity,click traffic trends,authority,and the time factors to correct the ranking value of web page. Then,on these basis,combining with user’s interest to conduct the secondary correction for the query results. It could make the web pages importance of weight calculation more reasonable.At last,the experiment was conducted by three groups including graduates,students and the teachers. The experimental results showed that,compared with conventional PageRank algorithm,the improved algorithm has higher precision,and better met the needs of users.更多还原