We present a kinetic model for opinion dynamics depending on the presence of social media contacts. Firstly, we describe how to model the evolution of the density of contacts starting from the microscopic level and, consequentially, deriving the kinetic model for the evolution of the connections. We use a dataset obtained from Twitter in order to select the set of parameters that allows to analytically simulate the current state of the connections on that social network. Secondly, from the binary interaction between two agents, we recover the Boltzmann-like equation that describes the evolution of the joint distribution of opinions and contacts. We perform a set of numerical simulations, showing how different possible users’ behaviors can be recovered changing the microscopic interaction between the agents. Finally, we show how to use sentiment analysis tools to obtain real distributions of opinions from Twitter, and we present a possible way of reconstructing the parameters of the interaction kernel to simulate the opinions evolution observed from the data. Joint work with G. Albi and G. Dimarco