Frugal Foodie: A Cost-Based Restaurant Clustering and Location-Based Recommendation System

Selecting a restaurant that aligns with both individual budget constraints and geographic preferences remains a challenging task in urban environments, where users are often overwhelmed by the abundance of choices. Traditional restaurant search engines offer limited filtering and personalization, lacking the intelligence to dynamically adapt to user context such as real-time location and spending ability. This project presents a novel system that combines unsupervised machine learning and real-time geolocation services to generate personalized restaurant recommendations. By applying K-Means clustering on a preprocessed Zomato dataset, restaurants are grouped into distinct budget categories, enabling precise matching based on user-defined financial preferences. Furthermore, the integration of Google’s Geocoding and Places APIs allows the system to pinpoint the user’s current location or interpret a manually entered area name to fetch nearby restaurant data within a customizable search radius. A Flask-based web application provides a user-friendly interface to input preferences and view results, including restaurant names, ratings, addresses, and map links. Clustering effectiveness is evaluated using the Silhouette Score, ensuring meaningful segmentation. This intelligent, adaptive, and location-aware recommendation system streamlines decision-making and significantly enhances user convenience and satisfaction.