Abstract:

This work used a fresh perspective on mood. We created a novel framework using a new classification of emotions called emotion activity, alongside deep learning to extract feelings about climate change. As a result, we generated appropriate mood-targeting techniques. The aim of this work was to examine public perceptions about climate change following the release of Netflix's film Don't Look Up, and to hypothesise messaging for a campaign that can address these public perceptions efficiently and appropriately. We analysed tweets about Don't Look Up using a novel data science methodology that included classification of emotions as active and passive using deep learning. We found that the film sparked active emotions, such as fear and panicking about climate change, and polarisation of opinions about climate change. However, we also detected that a concrete action plan is missing. People would like to do more about climate change, but are not sure what that should be, ultimately causing further feelings of despair. These findings served as the basis for mood targeting based on the science of nudging. We suggest that the most effective campaign should make use of cognitive biases and moods of target audiences, such as the incremental foot-in the door technique with exclamation marks and warm, natural colours for happy, relaxed people, or a self-efficacy nudge with dark imagery and short sentences for unhappy, depressed people.