摘要
针对传统的协同过滤算法推荐准确率较低的问题,提出一种基于信任社区的个性化推荐策略.首先利用社区发现算法,从用户网络中挖掘出具备类似兴趣喜好的信任社区,然后实施基于社区的个性化推荐.为及时发现用户的兴趣迁移及恶意攻击节点,引入一种信任度的反馈评价机制,将交易后的评价数据与预期值进行比较,以实现信任度的自适应更新.实验数据显示,该算法使得系统推荐准确率得到有效提高,从而提高了用户对系统的信任度.
According to the traditional collaborative filtering recommendation low accuracy problem,proposes a strategy of community based personalized recommendation trust. Firstly,community discovery algorithm,which has similar interests of the mining community trust from the users in the network,then the implementation of personalized recommendation based on community. In order to find out the interest migration and malicious node for users,the introduction of a trust evaluation feedback mechanism,the evaluation data and the expected transaction values were compared,in order to achieve adaptive trust update. The experimental results show that the proposed algorithm can effectively improve the accuracy of recommendation and enhance the user experience.
作者
马小琴
杨利
MA Xiao-qin;YANG Li(College of Mathematics and Computer Science,Chi Chou University,Chi Zhou 247000,Anhui,China)
出处
《山西师范大学学报（自然科学版）》
2018年第3期39-43,共5页
Journal of Shanxi Normal University(Natural Science Edition)
基金
安徽省教育厅自然科学研究重点项目(KJ2015A264)
池州学院校级自然科学研究重点项目(2016ZRZ010
2014ZRZ008)
关键词
协同过滤
推荐准确率
兴趣迁移
信任社区
Collaborative filtering
recommendation accuracy
interest migration
community trust
