Social Media Attitudes about Mining for the Green Transition in Europe Using Machine Learning Techniques

The Euorpean Union's Horizon program has recently funded the SEMACRET project for the sustainable exploration of critical raw materials. This is of particular importance as the EU seeks to increase its energy and mineral self-sufficiency and decrease its dependence on an external and potentially volitile supply chain. Among the technical challenges of novel resource identification and development, there are also many social aspects of exploration that must be understood and appreciated if the social license to explore is to be gained and resource exploration projects are to move forward. Understanding stakeholder perspectives, concerns, priorities, and values is crucial to developing policies and programs that will result in the accomplishment of these goals. That is why SEMACRET has a working package dedicated to exploring these facets of resource development within member states, local communities, and in social media. In particular, attitudes expressed on social media can be difficult to understand due to the volume of information, the ambiguous status of users as stakeholders, and the semi-anonymous nature of social media interactions. To address these challenges, researchers from SEMACRET's social science working package have worked to develop a machine learning application that uses natural language processing techniques to identify, differentiate, and understand perspectives on local mineral exploration expressed on social media. This presentation explains the methodology (latent Dirichlet allocation) and shows results from the four EU member states (Poland, Portugal, Czech Republic and Finland) that are the focus of SEMACRET's exploration research.