Learning to Deceive Knowledge Graph Augmented Models via Targeted Perturbation

This is a PyTorch implementation of our ICLR 2021 paper:

Learning to Deceive Knowledge Graph Augmented Models via Targeted Perturbation
Mrigank Raman, Aaron Chan*, Siddhant Agarwal*, Peifeng Wang, Hansen Wang, Sungchul Kim, Ryan Rossi, Handong Zhao, Nedim Lipka, Xiang Ren.
ICLR 2021.
*=equal contritbution

Note: This repo is still under construction. Please check again later!

Getting Started

Clone the repository

git clone https://github.com/INK-USC/deceive-KG-models.git

Question Answering

1. Download the pretrained models and datasets

cd deceive-KG-models/obqa

bash scripts/download.sh

cd data/cpnet

wget https://csr.s3-us-west-1.amazonaws.com/tzw.ent.npy

2. Preprocess the data

cd deceive-KG-models/obqa

python preprocess.py

By default, all available CPU cores will be used for multi-processing in order to speed up the process. Alternatively, you can use -p to specify the number of processes to use:

python preprocess.py -p 20

3. Train the base classifier

Then train a grn model (for example)using the command:

python grn.py -ds obqa --encoder bert-base-uncased -bs 64 -mbs 4 -dlr 1e-3

Similarly train RN and GN as well.

4. Train the triple classifier

python Get_neg_triples.py

python deep_triple_classifier.py

5. Pruning the Graph for only the useful nodes

python new_graph.py

6. Running Heuristics

python heuristics.py

The attributes are:

-np --num_pert: number of perturbations

--type type of perturbation(rel for Relation Swapping, edge for Edge Deletion and edge1 for Edge Rewiring)

For Relation Replacement use the command:

python train.py --mode_type eval --num_epochs 1 --save_dir ./saved_models/KG/model_25 --model_id 5 --enable_shuffle --dqn_lstm_len 100 --dqn_batch_size 16 --dqn_train_step 50 --log_path log_25.csv --steps_after_collecting_data 2000

7. Training the RL agent

python train.py --mode_type train --num_epochs 1 --save_dir ./saved_models/KG/model_25 --model_id 1 --enable_shuffle --dqn_lstm_len 100 --dqn_batch_size 16 --dqn_train_step 50 --log_path log_25.csv --steps_after_collecting_data 2000

8. Evaluating the RL agent

python train.py --mode_type eval --num_epochs 71801 --num_steps 70000 --save_dir ./saved_models/KG/model_25 --model_id 1 --debug_mode

what you have to change for specific model:

change of GPU number

change saved path to coincide with your saved model: ./saved_models/KG/model_25;

change num_steps: the number of steps you want to perturb, 70000 as default

Recommendation based experiments

cd deceive-KG-models/RippleNet

1. Train the base classifier

bash scripts/run_movie.sh

or

bash scripts/run_music.sh

2. Run Heurisitics

python main_RN.py

The attributes are: --dataset: movie or music

--type : Type of perturbation

--num_pert : Number of perturbations

3. Train RL agent

python train.py --mode_type train --num_epochs 1 --save_dir ./saved_models/KG/model_25 --model_id 1 --enable_shuffle --dqn_lstm_len 100 --dqn_batch_size 16 --dqn_train_step 50 --log_path log_25.csv --steps_after_collecting_data 2000

4. Evaluate the RL agent

python train.py --mode_type eval --num_epochs 71801 --num_steps 70000 --save_dir ./saved_models/KG/model_25 --model_id 1 --debug_mode