Data Poisoning Attacks on Graph Convolutional Matrix Completion

Zhou, Qinwei; Ren, Yi; Xia, Tianyu; Yuan, Lifeng; Chen, Linqiang

doi:10.1007/978-3-030-38961-1_38

Cited by 8 publications

(13 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For RQ1, we use average promote score (APS) to measure the effectiveness of different attacks. It changes from the indicator average promotion ratio in literature [27], and it can reflect the macro attacking effect. The evaluation defined as below.…”

Section: Metricsmentioning

confidence: 98%

“…We adopt two widely used real-world datasets in recommend system, which are named ML-100K [27] and Yelp [13]. ML-100K.…”

Section: Datasetmentioning

confidence: 99%

“…A popular attack be modified by the method from literature [27]. We choose the top n − 1 items from the training data popularity as fillers.…”

Section: Popular Attack Based Item (Pa-i)mentioning

confidence: 99%

“…A variant of a popular attack in literature [27]. The fake score setting is similar as the PA-I, and the strategy of choosing filling users is different.…”

Section: Popular Attack Based Score (Pa-s)mentioning

confidence: 99%

“…Average promotion ratio (APR). Average promote ratio (APR) [27] is a more fine-grained metric. Using it, we can distinguish the improvement effect of the model on different target products and can make the more targeted analysis.…”

Section: Metricsmentioning

confidence: 99%

See 4 more Smart Citations

Semantic Shilling Attack against Heterogeneous Information Network Based Recommend Systems

Ren

Yuan

et al. 2022

IEICE Trans. Inf. & Syst.

Self Cite

View full text Add to dashboard Cite

The recommend system has been widely used in many web application areas such as e-commerce services. With the development of the recommend system, the HIN modeling method replaces the traditional bipartite graph modeling method to represent the recommend system. But several studies have already showed that recommend system is vulnerable to shilling attack (injecting attack). However, the effectiveness of how traditional shilling attack has rarely been studied directly in the HIN model. Moreover, no study has focused on how to enhance shilling attacks against HIN recommend system by using the high-level semantic information. This work analyzes the relationship between the high-level semantic information and the attacking effects in HIN recommend system. This work proves that attack results are proportional to the high-level semantic information. Therefore, we propose a heuristic attack method based on high-level semantic information, named Semantic Shilling Attack (SSA) on a HIN recommend system (HERec). This method injects a specific score into each selected item related to the target in semantics. It ensures transmitting the misleading information towards target items and normal users, and attempts to interfere with the effect of the recommend system. The experiment is dependent on two real-world datasets, and proves that the attacking effect is positively correlate with the number of meta-paths. The result shows that our method is more effective when compared with existing baseline algorithms.

show abstract

Section: Metricsmentioning

confidence: 98%

“…We adopt two widely used real-world datasets in recommend system, which are named ML-100K [27] and Yelp [13]. ML-100K.…”

Section: Datasetmentioning

confidence: 99%

“…A popular attack be modified by the method from literature [27]. We choose the top n − 1 items from the training data popularity as fillers.…”

Section: Popular Attack Based Item (Pa-i)mentioning

confidence: 99%

“…A variant of a popular attack in literature [27]. The fake score setting is similar as the PA-I, and the strategy of choosing filling users is different.…”

Section: Popular Attack Based Score (Pa-s)mentioning

confidence: 99%

Section: Metricsmentioning

confidence: 99%

See 3 more Smart Citations