Sifting French Tweets to Investigate the Impact of Covid-19 in Triggering Intense Anxiety

Sifting French Tweets to Investigate the Impact of Covid-19 in Triggering Intense Anxiety Mohamed-Amine Romdhane, Elena Cabrio and Serena Villata Université Côte d’Azur, Inria, CNRS, I3S, France mohamed-amine.romdhane@etu.univ-cotedazur.fr, elena.cabrio@univ-cotedazur.fr, villata@i3s.unice.fr RÉSUMÉ Les réseaux sociaux peuvent être exploités pour comprendre les sentiments et les émotions des personnes en temps réel et cibler les messages de santé publique en fonction de l’intérêt et des émotions des utilisateurs. Dans cet article, nous étudions l’impact de la pandémie COVID-19 dans le déclenchement des crises d’angoisse, en nous appuyant sur les messages échangés sur Twitter. Plus précisément, nous fournissons : i) une analyse quantitative et qualitative d’un corpus de tweets en français liés au coronavirus, et ii) une approche en pipeline (un mécanisme de filtrage suivi par des méthodes de réseaux de neurones) pour classer de manière satisfaisante les messages exprimant de l’anxiété sur les médias sociaux, en considérant le rôle joué par les émotions. ABSTRACT Sifting French Tweets to Investigate the Impact of Covid-19 in Triggering Intense Anxiety. Social media can be leveraged to understand public sentiment and feelings in real-time, and target public health messages based on user interests and emotions. In this paper, we investigate the impact of the COVID-19 pandemic in triggering intense anxiety, relying on messages exchanged on Twitter. More specifically, we provide : i) a quantitative and qualitative analysis of a corpus of tweets in French related to coronavirus, and ii) a pipeline approach (a filtering mechanism followed by Neural Network methods) to satisfactory classify messages expressing intense anxiety on social media, considering the role played by emotions. MOTS-CLÉS : détection de l’anxiété, COVID-19, données Twitter, apprentissage automatique, apprentissage profond. KEYWORDS: intense anxiety detection, COVID-19, Twitter data, machine learning, deep learning. 1 Introduction The COVID-19 pandemic - that overtook most of the world’s countries - is forcing government- issued lockdowns, strict hygiene regulations and is ultimately causing global panic, uncertainty and fear. During this crisis, social media are representing a valuable source of news and a medium for expressing people’s opinions and sentiment about the emergency 1 and the restrictive measures deployed by the different countries to fight COVID-19 spread. Given that social media are proving instrumental in keeping people connected during the crisis, they turned out to be beneficial for mental 1. https://www.medrxiv.org/content/10.1101/2020.04.03.20052936v1 Actes de la 28e Conférence sur le Traitement Automatique des Langues Naturelles Lille, France, 28 juin au 2 juillet 2021 Volume 1 : conférence principale, pages 219–226. Cette œuvre est mise à disposition sous licence Attribution 4.0 International. health, in helping to combat widespread feelings of loneliness stemming from extended periods of isolation and social distancing. However, panic attacks are reported to be increasing during COVID-19, as people are increasingly worried about their health 2. Panic attacks are characterized by an intense fear and sense of feeling overwhelmed, according to the National Institutes of Mental Health. Fear of contamination, sleep disturbance and probability of an economic slowdown with potential job losses are among the major factors leading to depression and anxiety among people (as emerges …

Content truncated to 500 words