摘要
分析对象不满足彼此独立的假设,阻碍了社会网络推荐工作的开展。据此提出了基于活动事件的社会网络推荐模型,给出了该模型的组成结构和工作流程,以及实现关键;该模型将用户活动事件作为推荐工作的关键依据与切入点,通过事件监测与建模,最终用基于子网的匹配算法实现推荐列表生成,通过过滤实现高精确度。仿真实验证明,该模型具有较高的处理效率和推荐精度。
Analyzed objects don＇t fit to the conditional independence assumption, which hinders the de-velopment of SNS recommendation systems. In order to deal with them, a novel system was proposed withclient action event recognition. And its model structures, working flows and key technologies were givenas following. The model utilized client action events to generate the key nodes of recommendation, andevent monitoring model to make a sub-net algorithm to match nodes. Simulation results show that the mod-el has better processing performance and recommendation accuracy than others do.
出处
《情报科学》
CSSCI
北大核心
2015年第11期140-144,共5页
Information Science
基金
教育部人文社会科学研究项目(10YJCZH169)
福建省社会科学研究项目(2010B064)
华侨大学科研基金资助项目(07HSK02)
关键词
信息处理
社会网络
推荐模型
活动事件
匹配
子网
information processing
social networking services
recommendation model
action event
match
sub-net
