FREx (the Framework for Recommendations with Explanations) is a Python package to support the development of recommender systems that use rule-based or knowledge-based steps in the recommendation process.
FREx is developed for Python 3.8+
You can use FREx by installing it via pip
$ pip install git+https://github.com/solashirai/FREx@master#egg=frexNote that one of the requirements, ortools, might require you to be using a 64-bit installation of Python. If you see installation errors related to ortools, consider swapping to a 64-bit Python (if you were using 32-bit) or checking that your pip is upgraded.
You will now be able to use FREx in your project like any other Python package, with import frex.
More documentation about the project is available here.
The main classes and modules can be found in the frex/ directory. A toy example demonstrating the use of frex can be seen in the examples/ramen_rec directory.
Documentation generated using Sphinx can be found at https://solashirai.github.io/FREx.