Operators of military unmanned aerial vehicles (UAVs) work in highly dynamic environments. They have to complete numerous tasks, sometimes simultaneously, while maintaining high situational awareness (SA) and making rapid decisions. Their main focus is on mission management via the UAV's payload, yet, they continuously interact with the command and control (C2) map to obtain SA and make decisions. C2 maps, shared among forces in the environment, are cluttered and overloaded with information. We aim to develop a map display machine-learning based spatial-temporal algorithm that will identify the most relevant information items to the UAV operator and deliver the right visualized information on the C2 map at the right timing. Towards the algorithm development, experiments for collecting user-based importance data were conducted and analysed. For this, a designated UAV C2 Experimental System (UCES) has been developed. Results show high feasibility for the prediction model, allowing to move forward with the following steps of the algorithm development.