Recipes Recommendation System Based on Ingredient Availability and User Preferences

Cooking is an important and enjoyable activity that allows individuals to present delicious dishes. However, finding recipes that match the available ingredients and personal preferences can be a challenge. To address this, an introduced recommendation system utilizes a content-based filtering approach with Jaccard similarity to leverage the ingredients owned by users. Additionally, an item-based collaborative filtering approach using cosine similarity is employed to suggest recipes based on user preferences and those of other users. This approach yields accurate and relevant recipe recommendations, maximizing ingredient utilization and enhancing the cooking experience. The optimal solution for ingredient-based recommendations is achieved by combining ingredient ownership percentage and Jaccard similarity. Evaluation of recommendations based on user preferences also reveals varying outcomes, with an accuracy rate reaching 70.76%.