batmanscode / ruler

Web app to generate recommendations using association rule learning.

Ruler 📏

This web app uses machine learning to generate recommendations (association rule learning) from an input CSV.

The input CSV needs the following input columns:

• Unique transaction ID (like an invoice number)
• Item name

To see what the output looks like, click 'Get rules' and rules will be generated using built in sample data.
Sample data used is a variation of: Online Retail Data Set

Demo

How to Run Locally

Clone this repository into a new environment, install dependencies from requirements.txt and enter the following in your terminal:

streamlit run ruler.py

This will create a local web server which should open in your default browser. If not, just use one of the links returned in your terminal.

Additional Info

• Web app built with Streamlit.
• Associaton rules generated using Pycaret's Association Rules module.