Intelligent cross-platform product comparison using LSTM-based sentiment analysis and fuzzy logic ranking for E-Commerce

This project introduces a cross-platform product comparison system that is AI-powered and allows users to access an integrated product comparison system across a variety of e-commerce websites. The system will take product URLs of various online market places and will automatically scrape detailed product information, product name, price, star rating, technical specifications and customer reviews, with powerful web scraping models. Intelligent evaluation requires the extracted reviews of customers to be fed into a Bidirectional Long Short-Term Memory (Bi-LSTM) deep learning sentiment-classifying model with a positive, neutral, and negative classification. The results of the sentiment analysis are also combined with other quantitative variables that include the price competitiveness, rating distribution and review volume. To obtain an understandable, transparent and reliable decision support, a fuzzy logic decision module combines these multi-dimensional inputs to produce AI-based recommendations and confidence scores. All the processed data as structured product metadata, sentiment summaries, fuzzy evaluation scores, and final recommendation outputs are stored safely in a Supabase cloud database, and it is easy to retrieve them and store them in large volumes and compare them in real-time. The system is intended to be a multi-platform solution running on web platforms to provide a consistent, data-driven, and user-centric experience, comparing products, which is driven by real-time e-commerce data.