Adversarial Recommender Systems: Attack, Defense, and Advances

Anelli, Vito Walter; Deldjoo, Yashar; DiNoia, Tommaso; Merra, Felice Antonio

doi:10.1007/978-1-0716-2197-4_9

Cited by 8 publications

(2 citation statements)

References 72 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These ML-based attacks can be divided into two classes: evasion attack and poisoning attack, based on attack timing [53,314]. In this part, we introduce two types of ML-based attacks and Adversarial Machine Learning (AML) methods to defend against these attacks.…”

Section: Machine-learned Adversarial Attacksmentioning

confidence: 99%

See 1 more Smart Citation

Towards trustworthy recommenders: building explainable and unbiased recommendation systems

View full text Add to dashboard Cite

show abstract

Section: Machine-learned Adversarial Attacksmentioning

confidence: 99%

“…In recommendation tasks, most attackers aim to adjust the representations of users or items to add noise [314]. For example, He et al [316] introduced a method to add perturbations to parameters of embedding layers based on the Fast Gradient Sign Method (FGSM) [317], which has been widely applied to generate adversarial perturbations.…”

Section: Evasion Attackmentioning

confidence: 99%