convmf

code for "Convolution Matrix Factorization for Document Context-Aware Reccomendation"

Setup

install python module.

pip install -r requirements.txt

download rating data

./download_data.sh

download movie description data
- visit https://www.kaggle.com/tmdb/tmdb-movie-metadata
- download data and expand zip into ./data/
- data directory must be as follows:
```
  - data/
    - ml-10M100K/
    - tmdb-5000-movie-dataset/
```

Run

python preprocessing.py

python train.py