RytTnk / PACE2017

A step-by-step Keras implementation of PACE (Preference And Context Embedding) described in our KDD 2017 paper.

PACE

A step-by-step Keras implementation of PACE (Preference And Context Embedding) described in our KDD 2017 paper. To run the code, you need to have Python 3 and iPython Notebook installed.

Please cite the following work.

Carl Yang, Lanxiao Bai, Chao Zhang, Quan Yuan and Jiawei Han. 2017. Bridging Collaborative Filtering and Semi-Supervised Learning: A Neural Approach for POI Recommendation. In Proceedings of KDD ?17, Halifax, NS, Canada, August 13-17, 2017, 10 pages.

Usage:

• Use bash download_data.sh to download the Gowalla data or visit Yelp to download the Yelp data.
• Run python3 dataset.py for data preprocessing (slight modifications needed to match specific data formats).
• Start iPython Notebook Server ipython3 notebook
• Sequentially run cells in train.ipynb

If you are using remote machine, you can:

• Start iPython Notebook Server on remote machine: ipython notebook --no-browser --port=8889
• Redirect ssh connection to localhost ssh -N -f -L localhost:8880:localhost:8889 <user>@<host>
• Open browser and go to <user>@<host>:8880