https://github.com/hsm207/fair-recsys
Code to implement a reranking method to improve recommendation diversity
https://github.com/hsm207/fair-recsys
Last synced: 7 months ago
JSON representation
Code to implement a reranking method to improve recommendation diversity
- Host: GitHub
- URL: https://github.com/hsm207/fair-recsys
- Owner: hsm207
- License: gpl-3.0
- Created: 2019-12-14T11:15:22.000Z (almost 6 years ago)
- Default Branch: master
- Last Pushed: 2019-12-14T16:42:04.000Z (almost 6 years ago)
- Last Synced: 2025-03-25T10:51:15.400Z (8 months ago)
- Language: Jupyter Notebook
- Homepage:
- Size: 54.7 KB
- Stars: 4
- Watchers: 2
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Introduction
This repository contains code to accompany my Medium blog post titled [A Simple Post-Processing Step To Improve The Fairness Of Collaborative Recommender Systems](https://medium.com/@_init_/a-simple-post-processing-step-to-improve-the-fairness-of-collaborative-recommender-systems-4cfbdb47c4ea).
# Prerequistites
The code was developed using Anaconda Python 3.7.
The calls to PySpark are executed using Databricks's [dbconnect](https://docs.databricks.com/dev-tools/databricks-connect.html) tool.
# Usage
The main results are in the [notebooks](./notebooks) folder. Execute them sequentially. The files ending with `.py` instead of `.ipynb` means that
those are Databricks notebooks.
The code that implements the reranking function is in the [preference_reranker.py](./src/preference_reranker.py) file.
# Contributing
Feel free to raise an Issue if you have any questions, feedback or comments.