원문정보
초록
영어
Aiming at the low limitation of shilling attack detection technology unsupervised degree, this paper takes the group effect attack profile as the breakthrough point to construct the attack profile groups and the corresponding genetic optimization objective function of quantitative measure of the effects, and prove that the maximum value of the objective function in the ideal state marks the optimum detection effects in ideal situation. On this basis, the combination of genetic optimization process will be adaptive parameter posterior inference and objective function, and proposes the Iterative Bayesian Inference Genetic Detection Algorithm (IBIGDA).Experimental results show that IBIGDA can effectively detect shilling attacks of typical types even in lack of the system or attack-related prior parameters. IBIGDA algorithm can detect common shilling attack, unsupervised degree is high, with the actual application requirements.
목차
1. Introduction
2. Iterative Bayesian Inference Genetic Detection Algorithm (IBIGDA)
2.1. The Statistical Characteristics of Exists Attack between the User Profile Attack
2.2. Generalized Variance Induced Attack Profile Group Effect Metric
2.3. Genetic Optimization Objective Function
2.4. Algorithm Description and Interpretation
3. Experiment Design and Discussion
4.1. Data Sets and Experimental Setup
4.2. The Example Analysis of Shilling Attack Detection Process
4.3. Detection Effect of Shilling Attack
5. Conclusion
References