Skip to content

Hacxmr/Eco-Friendly-Tribe

Repository files navigation

Eco-Friendly Fabric Recommendation System

This project is a fabric recommendation system that helps users find eco-friendly and sustainable fabric options by suggesting similar fabrics based on an initial selection. The recommendation system is powered by FAISS (Facebook AI Similarity Search), and it uses Streamlit for an interactive frontend experience.

This system is a part of the Eco-Friendly Tribe project, which aims to promote sustainable fabric choices.

Eco-Friendly Fabric Recommendation System

Features

  • Fabric Selection: Users can choose a fabric from a dropdown menu to find similar fabrics.
  • Text Vectorization: Uses TF-IDF to transform fabric titles and descriptions into numerical vectors.
  • Label Encoding: One-hot encodes categorical labels.
  • Similarity Search: Utilizes FAISS to find and recommend fabrics similar to the selected one.

Dataset

The dataset used is FabricFrontiers, a CSV file with columns like idx, title, description, source, and label.

Installation

  1. Clone the repository:

    git clone https://github.com/Hacxmr/Eco-Friendly-Tribe.git
    cd Eco-Friendly-Tribe
  2. Install the required dependencies:

    pip install streamlit pandas faiss-cpu scikit-learn numpy

Usage

Run the Streamlit app:

streamlit run streamlit_app.py