Social media platforms can be a source of recommendations for various items or products. Digital Marketing takes advantage of this potential, especially when using influencers to evaluate their items or products on social media platforms. Each influencer can be a recommender of items or products. Thus, one can question why this does not occur similarly in the area of science. This work presents a study in which shares/mentions were collected on a social media platform, Twitter, about scientific documents (for example, journal or conference papers, thesis, reports, and patents). We pretend to obtain the polarity of these shares/mentions (positive, negative, and neutral) is obtained. It was found that most of the shares/mentions are neutral and few are considered positive, which leads to the conclusion that there are shares/mentions about science and that, despite the lack of detected polarity they can serve as a basis for a scientific recommendation.