Pandemic emergency: A multi-level AI framework for analyzing psychological impact through public discourse

Area of Science:

• Psychology
• Public Health
• Sociology

Background:

• The Covid-19 pandemic created global uncertainty, impacting health, social, economic, and environmental stability.
• Psychology is reassessing its role in supporting change during emergencies.
• Qualitative approaches can address population needs during crises.

Purpose of the Study:

• To explore Italian social media discourse during the Covid-19 pandemic (April 2020-May 2021).
• To observe the pandemic's impact on public health.
• To identify clinical implications from population-expressed concerns.

Main Methods:

• Text analysis of social media productions from the Italian population.
• Annotation system with 24 discursive modalities (human and machine learning).
• Content detection and descending hierarchical classification using IRaMuTeQ.

Main Results:

• Discourse analysis revealed a need for clinical and sociopolitical support.
• Individuals' discourse often focused on the "here and now," impeding future health goals.
• Prioritizing institutional critique over personal accountability led to passive observation rather than proactive well-being management.

Conclusions:

• Machine learning models can enhance clinicians' understanding of population health needs during emergencies.
• AI tools offer greater precision and speed for operational decision-making in crisis management.
• Interventions should address both individual accountability and sociopolitical factors for improved public health outcomes.