Product Recommendation using Sentiment Analysis for E-Commerce (PRSA)

PRSA is an advanced product recommendation system designed for e-commerce platforms. It includes three main modules:

1. Daraz Scraper: A web scraper built using Scrapy and Selenium to extract product URLs, details, and reviews from Daraz, storing the data in PostgreSQL.
2. Sentiment Analysis: Utilizes a fine-tuned DistilBERT model trained on Amazon datasets to perform sentiment analysis, integrated into a Flask API for seamless recommendations.
3. Website: A full-stack application with a Node.js backend and React frontend, implementing JWT authentication. Users can view products and recommendations, while admins can analyze product reviews and manage the recommendation engine. Recommendations are generated using a weighted score of sentiment (60%), rating (30%), and price (10%).

This project demonstrates expertise in web scraping, natural language processing, API development, and full-stack web technologies.