Ανάπτυξη Android εφαρμογής προτάσεων συνταγών βάση των υλικών του χρήστη αξιοποιώντας το Spoonacular API και μεγάλα γλωσσικά μοντέλα (LLMs) - FridgeChef
Development of Android recipe recommendation App based on user’s ingredients powered by the Spoonacular API and Large Language Models (LLMs) - FridgeChef

View/ Open
Keywords
Συνταγές ; LLMs ; Spoonacular API ; Pantry ; FirebaseAbstract
This undergraduate thesis focuses on the development of an Android application, specifically one that suggests recipes based on the ingredients the user already has at home. The main goal of the application is to simplify the daily routine of individuals who are not specialized in cooking, while also helping to reduce food waste by utilizing ingredients that might otherwise go unused. The application uses the Spoonacular API to recommend appropriate recipes, based on the ingredients stored in the user's personal digital pantry. Users can view detailed step-by-step instructions for each recipe, share them via social media platforms, save them for future use, and gain insights into what ingredients are commonly used by other users of the app.
In addition, the application incorporates the Nutri-Score food rating system to provide an assessment of the nutritional quality of the user's pantry contents. Leveraging Large Language Models (LLMs), the app enables ingredient recognition using the device's camera, allowing users to add items directly to their digital pantry. It also provides nutritional analysis for each recipe, offering users a comprehensive breakdown of the nutrients in the dishes they plan to prepare. All user data, including stored ingredients and saved recipes, is securely stored in Firebase. The application combines modern technologies to deliver a personalized, intelligent, and health-conscious cooking experience.