In this seminar, we will discuss how scientists have built mathematical models of opinions spreading in large populations. These models describe the spreading of information on (social) networks composed of nodes — here the nodes symbolise individuals — who can influence each other when connected by a network edge — testifying to the existence of a social link through which information can flow. It is important to observe that these mathematical models are based on major simplifications which do not allow us to describe with precision the real world and our complex society and in particular cannot predict the behaviour of a given individual. Therefore, their predictive power is somehow limited to generic individuals and they cannot anticipate with precision the outcome of future events. Nevertheless, these models are important to gain useful insights into what factors facilitate or hinder opinion spreading, and into what conditions one outcome is more probable than others. They also offer the possibility to study potential scenarios, otherwise impossible to realise for ethical and practical issues. These models thus often speak in terms of probabilities; predictions are never certain because uncertainty is an intrinsic factor of processes involving human behaviour and large societies. Building a model can be done starting from scientists’ intuition and assumptions or, alternatively, using (big) data. Both approaches come with flaws and limitations; our seminar will highlight the importance of considering these limitations when drawing conclusions from a mathematical model. Reservation mandatory >>>
The mathematical diffusion of opinions in social networks
Location: Rue de Livourne, 38, 1000 Brussels
