Skip to content

Latest commit

 

History

History
41 lines (26 loc) · 1.26 KB

README.md

File metadata and controls

41 lines (26 loc) · 1.26 KB
Raw

Collaborative Memory Network for Recommendation Systems

Ebesu, Shen, Fang - The 41st International ACM SIGIR Conference on Research & Development in Information Retrieval - SIGIR '18

Link to paper.

This repository by Aditya Srivastava is a PyTorch port to the original TensorFlow project.

Running the Collaborative Memory Network

  1. Start a Jupyter server and open the notebook.
  2. Run notebook cells, beginning at the top and moving downwards.
  3. Pretraining the network is optional; pretrained embeddings have been provided already in pretrain/.

Requirements

  • Python 3.6.7
  • Jupyter
  • torch 1.1.0
  • numpy 1.16.4
  • tqdm 4.32.2

Data Format

The structure of the data in the npz file is as follows:

train_data = [[user id, item id], ...]
test_data = {userid: (pos_id, [neg_id1, neg_id2, ...]), ...}

Acknowledgements

My thanks to Travis Ebesu, one of the authors of the paper linked above, for helping debug the project and clear up any questions I had along the way.

Issues

Please feel free to contact me/raise an issue in case of any questions/bugs.