This approach mainly has 5 modules, preprocessing and kg creation, generating kg embedding, training reinforcement learning agent, recommending items for customers, and path generation and evaluation.
Run the below three .ipynb files in the same order as they are listed.
- Preprocess the data:
run the processing.ipynb file
- KG creation, embeddings generation, and model building:
run the KG_RL_Rec_Sys.ipynb file
- Evaluation and Path generation:
run evaluation.ipynb
-
Python >= 3.6
-
PyTorch = 1.0
Install Python packages using provided requirements.txt file.