FastAPI Brand-Company Matcher

Overview

This project is a FastAPI application designed for matching brand and company names using fuzzy string matching. It leverages the rapidfuzz library for efficient string matching and integrates with a JSON dataset to provide accurate and fast brand-company associations.

Features

• Fuzzy String Matching: Utilizes rapidfuzz for matching user-input brand and company names to a dataset.
• FastAPI Framework: Built using FastAPI for efficient and easy-to-use web API development.
• Dynamic Dataset Integration: Integrates with a JSON-based dataset for brands and companies, allowing dynamic data handling.
• Customizable Thresholds: Utilizes environment variables for customizable matching thresholds.

Installation and Setup

1. Clone the Repository:
   • Clone the repository from GitHub.
2. Install Dependencies:
   • Install required Python libraries: fastapi, numpy, pandas, rapidfuzz, uvicorn.
3. Dataset Preparation:
   • Place your DatasetMapping.json file in the project directory.
4. Environment Variables:
   • Set THRESHOLD_COMPANY_MATCH and THRESHOLD_CONF_LEVEL in your environment for customization.

Running the Application

• Run the application using Uvicorn: uvicorn main:app --reload
• The application will be served at http://127.0.0.1:8000/.

API Usage

• Endpoint: GET /
• Parameters:
  • brand: The brand name to match.
  • company: The company name to match.
• Returns: A JSON response with the best match and confidence level.

Contributing

Contributions to enhance or expand the project are welcome. Please fork the repository and submit a pull request with your changes.

Future Enhancements

• Adding more datasets for broader matching capabilities.
• Improving the matching algorithm for higher accuracy.
• Incorporating advanced data preprocessing techniques.