摘要
网络用户信任感知推荐的准确性设计是提高用户间的社交网络辅助信息信任度的重要依据。传统的网络用户信任感知推荐算法采用的是基于社交网络服务和用户评分的推荐系统,主观性较大,协同过滤效果不好。提出一种基于网络动态干扰监控的信任感知推荐算法设计新方法,设计自适应神经模糊系统网络动态干扰监测算法,构建基于向量空间模型的信任度评价指标体系结构,通过调整网络拓扑权重向量设置信任度周期响应加权变量自适应函数,有效降低迭代算法的运算成本,避免了自适应神经模糊系统网络动态干扰监测加权权重成固化状态,提高抗干扰性能。实验结果表明,算法能使社交网络感知推荐模型的预测误差减少,推荐可靠性优于传统方法。
The accuracy design of network user trust aware recommender is key to improve social network auxiliary informa-tion between the user trust degree. The network user perceived trust traditional recommendation algorithm is used in the recommendation system, social network services and user ratings based on subjectivity, collaborative filtering effect is not good. A new method of algorithm design of recommendation trust is propsoed based on aware network dynamic interference monitoring and adaptive neural fuzzy dynamic interference, the design of monitoring system network construction algorithm, vector space model of trust evaluation index system based on trust, set the periodic response variable weighted adaptive function to adjust the network topology by weight, reduce the iterative algorithm cost, it avoids the adaptive neural fuzzy net-work dynamic interference monitoring weight into curing condition, the anti jamming capability is improved. The experi-mental results show that it can reduce the prediction error, and make social network aware recommendation model is superi-or to the traditional method, the recommended reliability is better.
出处
《科技通报》
北大核心
2014年第10期76-78,共3页
Bulletin of Science and Technology
关键词
网络动态干扰
用户信任感知
推荐算法
network dynamic interference
user perceived trust
recommendation algorithm
