KDEGroup / PC-Attack

Source code for AAAI'23 paper "Practical Cross-System Shilling Attacks with Limited Access to Data".

PC-Attack

Source code for AAAI'23 paper "Practical Cross-System Shilling Attacks with Limited Access to Data" [arXiv preprint].

Environment

• Linux with Python ≥ 3.6
• PyTorch ≥ 1.4.0
• 0.5 > DGL ≥ 0.4.3

Use the command "conda env create -n env_pc_attack -f environment.yml" to copy the exact same environment.

Data

The datasets used in our experiments can be found in the data folder.

We use datasets that are widely used in previous work.

Command Line Parameters

run.py is the main entry of the program, it requires several parameters:

• dataset: the source dataset used in the experiment (Possible values: ''filmtrust'', ''automotive'', "yelp", ''ToolHome''. Default is "yelp").
• target-dataset: the target dataset used in the experiment (Possible values: ''filmtrust'', ''automotive'', "yelp", ''ToolHome''. Default is "filmtrust").
• target-item: id of the target item (Default is 5).
• epochs: training rounds.
• gpu: GPU id.

Examples

Please refer to run.sh for some running examples.

python run.py --dataset yelp --target-dataset filmtrust --target-item 5 --epochs 64 --gpu 2