摘要
通过对现有个性化搜索引擎排序算法的研究,提出了一个新的排序算法。该算法首先在不同粒度上多次使用SVD技术和k-means聚类技术,将用户浏览历史及其所包含的词在不同层次上进行文档聚类和词聚类,创建两棵加权兴趣树:文档类树和词类树。其中,树中每个节点的权值表示用户对该类文档或该类词的感兴趣程度。接着,利用朴素贝叶斯分类器对搜索引擎得到的网页进行文档分类和词分类,并根据分类结果进行网页评分。最后,将网页根据文档得分降序排列。实验表明该方法能为用户提供更为精确的个性化排序。
A new ranking method is proposed based on the research on ranking algorithm for personalized search engines. SVD and k-means clustering algorithm are used for several times into different granularities to create two weighted interest trees：a document class tree and a word class tree. Each node in the tree is weighted and the weight represents the degree of interest of the user for that type of document or word. Then Bayesian classification algorithm and a scoring algorithm are applied to calculate the score of pages received by search engine. Finally, the scored pages are ranked in descending order. Experiments show that a more accurate rank can be achieved.
出处
《太原科技大学学报》
2013年第3期175-180,共6页
Journal of Taiyuan University of Science and Technology
