基于时域背离特征分析的托攻击检测算法

Shilling attack detection algorithm based on characteristic of deviate from time-domain

为提高托攻击检测的完整性、准确性和项目推荐可信度,使用托攻击评分与正常评分的数据差异度、正反序TOP窗口和后发评分特征以及密集评分特征,提出了一种基于时域背离特征分析的托攻击检测算法实现托攻击检测。实验结果表明:本文算法的F值平均为81.2%,最高可达92.7%,能同时提升托攻击检测的准确率、覆盖率,减少训练数据量,降低计算复杂度,同时还能兼顾托攻击防御。

In order to improve the completeness and accuracy of the shilling attack detection and the credibility of various recommendation information,this paper proposed a shilling attack detection algorithm based on characteristic of deviate from time-domain to detect shilling attack using the data diversity between the attack score and the normal score and the positive and negative top window and the later scoring and the dense scoring. The empirical analysis showed that the F of 81. 2%,upped to 92. 7%. The algorithm not only increased attack detection accuracy and coverage,but also reduced the amount of training data and the computational complexity and defense shilling attack.