【摘要】 针对Google PageRank算法中存在的"平均分配原则"及网络链接结构所造成的"旧网页问题"、"主题漂移问题",提出一种改进的网页排序算法N-PageRank。该算法通过对搜索日志进行数据挖掘,捕捉用户与搜索引擎之间的交互过程,发现隐藏在用户搜索行为背后的用户兴趣和搜索规律,利用用户行为反馈模型,分析网络日志里用户的各项行为特点,改善了排序结果的准确率,保证了搜索引擎的返回结果正是用户所希望看到的网页。实验证明该算法有效地降低了网页排序时的客观因素的影响,充分考虑了用户对于网页质量的评价,所得到的排序结果更加能够满足用户的需求。更多还原

【Abstract】 Analyzes the core algorithm of google-PageRank algorithm.Pointed out that the "principle of equal distribution",Web link structure caused by the "old Web page problem" and "topic drift".To solve these problems,based on the analysis on the original algorithm,proposes an improved N-PageRank algorithm,which through digging the search logs and found the user’s interest and the law of search,feedback models using user behavior to analyze the user behavior characteristics in the search logs,to improve the accuracy of ranking results.Experimental results show: the algorithm reduces effectively the influence of objective factors during page sorting;give full consideration to the evaluation of the user for the website quality,the sort results obtained more able to meet the needs of users.更多还原