Impact of Feature selection on content-based recommendation system

As a variety of foods and recipes have an impact on our daily lives and our choices, more and more recipes are being served in many restaurants around the world to cater the needs of customers and satisfy their different tastes. In order for customers to be able to decide where to eat, we present the Restaurant Content-based Recommender System. It classifies restaurants, with priority of order, according to their numerous features. We use Content based recommender system because the traditional collaborative filtering algorithms doesn't show satisfying performance as regards data sparsity, and also when having a contextual data. In this article, we will study the impact of feature selection on the performance of content based recommender system and we will use and show the performance of another method for calculation similarity between items in order to move away from classical cosine method.