MODEL AND DATASET OF SESSION-BASED RECOMMENDATION PROBLEM

DATASETS

MODEL

  • NARM: A model using GRU
  • SR-GNN: First model using GNN
  • GCSAN: SR-GNN + Attention :D
  • TAGNN++: Seem like GCSAN + target attention (seem super slow training time)
  • Updating....

Run code:

Training:

Run train.py in the folder having the name of the model to train the model. Time training using GPU in time.txt file

Testing:

Run test.py in the folder having the name of the model

python SRGNN/test.py