A GIS-Based Emotion Detection Framework for Multi-Risk Analysis in Urban Settlements

The application of sentiment analysis approaches to information flows extracted from the social networks connected to particular critical periods generated by pandemic, climatic and extreme environmental phenomena allow the decision maker to detect the emotional states of citizens and to determine which areas are most at risk and require specific resilient adaptation interventions. Of particular relevance today is the need to analyze the multiple risks generated by extreme phenomena in urban settlements in order for the decision maker to identify which areas are most at risk and prepare resilient intervention plans with respect to all the phenomena analyzed. In recent years, the COVID 19 pandemic emergency has forced citizens to undergo specific restrictions to protect their health; to these were added critical issues due to the occurrence of extreme climatic or environmental phenomena. In order to monitor pandemic and climate/environmental multi-risks in urban settlements, we propose a GIS-based framework in which an emotion detection method is applied to determine the prevailing emotional categories in urban study areas during pandemic periods and in the presence of extreme climatic phenomena. The framework was tested on a study area based in the six districts of the city of Bologna (Italy) in order to detect, based on the emotions expressed on social channels, which were the most critical city neighborhoods in pandemic periods and in the presence of extreme heat wave climatic events. The results show that the proposed model can represent a valid tool to support decision makers in identifying the most critical urban areas in the presence of pandemic and climate/environmental multi-risks.