基于用户反馈特征聚合的网页排序算法  

PageRank algorithm based on aggregation of users feedback characteristics

作　　者：王冲[1] 王凯+[1] WANG Chong WANG Kai(School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin 541004, China)

机构地区：[1]桂林电子科技大学计算机与信息安全学院,广西桂林541004

出　　处：《计算机工程与设计》2017年第8期2020-2024,共5页Computer Engineering and Design

摘　　要：为改善传统网页排序算法在网页排序过程中存在的忽略用户兴趣和网页欺诈等不足,提出一种基于用户反馈特征聚合的网页排序算法AFPR。通过Web用户反馈特征聚合,将不同的用户反馈特征聚合为一类,综合考虑用户对网页的有效点击率、最近访问频率、网页有效浏览时间等因素,将聚类结果以网页用户兴趣反馈特征依赖因子形式融入到传统网页排序算法中。仿真实验表明,AFPR算法能够在一定程度上改善排序质量,提升用户信息检索的满意度。To solve the problems of traditional PageRank algorithm, such as ignoring users＇ interest, page fraud and so on, the aggregation of users feedback characteristics PageRank （AFPR） was proposed. According to the traditional PageRank algorithm, users aggregation of users feedback characteristics was taken into consideration. The different features of users were aggregated into a class. Considering user＇s effective click rate, recent access frequency and other factors, the clustering results which were processed into a form of factor and traditional PageRank worked together in the AFPR algorithm. AFPR can improve the page ranking quality to some extent.

关 键 词：网页排序 用户反馈 特征聚合 影响因子 信息检索 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术]