Restaurant recommender data transformer

This collection of scripts is for transforming data collected by ZHAW for building a restaurant recommender.

Usage

Download this repository and install the required packages with pip.

pip install -r requirements.txt

Make sure you are using python 3.7+

The script expects a dataset split into three sets.

  • User features
  • Restaurant features
  • Ratings

Per default the scripts looks for those files in the input directory. They have to be in a CSV format following the raw data structure defined in the data report. To start the transforming you can execute the following command.

python transform.py