Food recommendation systems have become increasingly popular in recent years. With the growth of online commerce and the availability of a wide variety of food products, it becomes difficult for consumers to make informed choices. This is where food recommendation systems come into play. The objective of this work is to propose a food recommendation system to help consumers make choices that are tailored to their individual needs. The proposed system provides personalized nutritional recommendations taking into account preferences, allergies, and the health status of the patients. To do this, we have exploited dishes and their ingredients using the Word Embedding technique, notably GloVe and binary representation. Then, we measured distances using three different methods, namely, Cosine similarity, Euclidean distance, and Manhattan distance. The results of various experiments show that our system was able to recommend relevant foods taking into account the allergies and health issues of each user.