The recommendation engines have today an increasing influence on our consumption on the Internet. However, the data available to make a recommendation varies according to the users, the means of the industry and the type of products. In this paper, we focus on one hand, on the performances of matrix factorization as well as deep learning models and on the other hand, the recommendation on big data. A comparative study between several state-of-the-art models was carried out with an applicative purpose linked to a tyre data industrial.