Ολοκληρωμένη εφαρμογή διατροφής : εξατομικευμένες προτάσεις γευμάτων και συνταγών
Nutrition application : personalized suggestions for meals and recipes
![Thumbnail](/xmlui/bitstream/handle/unipi/16695/Panousi_p19131.pdf.jpg?sequence=6&isAllowed=y)
View/ Open
Keywords
Ευφυής βοηθός ; Eξατομικευμένες διατροφικές προτάσεις ; Ανοιχτά δεδομένα ; Τεχνητή νοημοσύνηAbstract
This thesis concerns the development of an Intelligent Assistant for nutrition suggestions and the generation of personalized meal plans. An open data set was selected as the training data set, comprising over 700,000 recipes from accredited sources and in addition, user reviews regarding them. The data was refined, checked, and sorted according to the most popular ingredients, thereby eliminating recipes that did not align with the country's cultural norms. Subsequently, the application was developed in MATLAB and enables users to create, connect, enter and manage relevant data in a web multiuser environment. It then provides users with the ability to select available materials and receive the most popular recipes that can be produced with them. Furthermore, the application provides nutritional information and the ability to display images of the recipes. Additionally, it connects to Open AI's API in order to create a weekly meal plan, tailored precisely to the user's ingredients and suggestions. This connection allows the intelligent assistant to efficiently generate personalised nutrition plans.