Social media platforms can be a source of recommendations for various items or products. Digital Marketing takes advantage of this, especially when using influencers to evaluate their items or products on social media platforms. Each influencer can be a recommender of items or products. However, one can question why this does not occur similarly in the science area. This work presents a study based on comments shared on social media platforms about scientific documents (for example, journal or conference articles, thesis, reports, and patents). We pretend to obtain the polarity (positive, negative, and neutral) for each scientific comment. Most mentions are neutral and some considered positive, which allows us to conclude that there are shares of comments about science and that they can serve as the basis for a science recommendation system.