Le masque, figure polaire de la crise de la Covid-19 : une exploration par NLP du flux des conversations Twitter (février - mai 2020)

The Covid-19 pandemic that has hit the planet offers a spectacular case study in disaster management. In this literature, the participatory paradigm is fundamental: the mitigation of the impact of the disaster, the quality of the preparation and the resilience of the society, facilitate reconstruction, but depend on the participation of populations. Being able to observe and measure the mental health of populations (anxiety, confidence, hopes, etc.), identifying points of controversy and the content of the discourse, remain necessary to accompany measures to encourage this participation. Social media, and in particular Twitter, offer valuable resources for exploring this discourse. The main result is based on the identification of the centrality of the mask figure and aims to establish the importance of the phenomenon. We show this quantitatively, and explore the concept using NLP methods. The background is a major change in the understanding of the crisis. If, at the beginning of the cycle, it is perceived in an exotic way, it then becomes endemic to the social body. We exploit here a database of 2.1 million tweets extracted from a corpus of 110 million, elaborated by an international information science team and dealing with the variants of #Covid-19, #coronavirus, etc: the Covid-19 Twitter data set.