"Eat What" System

Location: cogs/eat/

The "Eat What" system is a sophisticated, personalized restaurant recommendation engine integrated into the bot. It combines web crawling, database storage, and machine learning to learn a server's food preferences and provide tailored suggestions. This feature is primarily accessed through the eat option in the /internet_search command.

Core Components

The system is divided into several key components:

• Database: Manages the storage of search history and user preferences.
• Google Maps Crawler: Fetches real-time restaurant data from Google Maps.
• Machine Learning Model: Trains on user feedback to provide personalized recommendations.
• UI (Embeds & Views): Provides the interactive Discord interface for users.

Workflow

1. A user initiates a search via /internet_search search_type: eat.
2. If the user provides a keyword, the Google Maps Crawler searches for that type of food.
3. If no keyword is provided, the Machine Learning Model predicts a food recommendation based on the server's past ratings.
4. The crawler fetches details for a randomly selected restaurant.
5. The UI displays the recommendation in an interactive embed with buttons (Map, Menu, Like, Dislike, etc.).
6. The search result is saved to the Database.
7. When a user clicks "Like" or "Dislike", their rating is saved to the database.
8. This new rating triggers the Machine Learning Model to retrain, refining its future recommendations for that server.