This repository contains a reference implementation of our Multi-Pointer Co-Attention Networks (MPCN) proposed in KDD 2018.

Preprint is available here: https://arxiv.org/abs/1801.09251

Note: This is a preliminary rough sketch of the MPCN implementation. The main functions are already in this repository. The next update will including running instructions and dataset prep. Since this is mainly copied from another repository, some dependencies might be broken (for now). I plan to test this repo in isolation soon. Stay tuned.

1. Tensorflow 1.7
2. Python 2.7

If you find our source code useful, please consider citing our work.

@article{DBLP:journals/corr/abs-1801-09251,
  author    = {Yi Tay and
               Luu Anh Tuan and
               Siu Cheung Hui},
  title     = {Multi-Pointer Co-Attention Networks for Recommendation},
  journal   = {CoRR},
  volume    = {abs/1801.09251},
  year      = {2018},
  url       = {http://arxiv.org/abs/1801.09251},
  archivePrefix = {arXiv},
  eprint    = {1801.09251},
  timestamp = {Fri, 02 Feb 2018 14:20:25 +0100},
  biburl    = {https://dblp.org/rec/bib/journals/corr/abs-1801-09251},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}