Monitoring the Emotional Response to the COVID-19 Pandemic Using Sentiment Analysis: A Case Study in Mexico

Area of Science:

• Computational Social Science
• Public Health
• Mental Health Analytics

Background:

• The COVID-19 pandemic significantly impacted global lifestyles, mental health, and public health programs.
• Traditional methods for assessing population mental health are often slow and costly, hindering timely policy adjustments.
• Social media platforms offer a vast, real-time data source for understanding public sentiment and well-being.

Purpose of the Study:

• To track and quantify sentiment shifts in the Mexican population during the COVID-19 pandemic.
• To analyze the relationship between public mood, pandemic evolution, news cycles, and governmental policies.
• To assess the utility of social media data for mental health monitoring.

Main Methods:

• Analysis of 760,064,879 public tweets from Mexico using open sentiment analysis tools.
• Examination of collective mood shifts in response to pandemic events and policies.
• Quantification of sentiment polarity scores and weekly seasonality.

Main Results:

• Consistent negative sentiment polarity scores (around -0.15) observed, with weekly seasonality.
• Increased public controversy noted following the termination of lockdown measures and during holiday periods.
• A 2.38-fold increase in social media usage was observed, indicating its use as a coping mechanism for social isolation.

Conclusions:

• Social media sentiment analysis provides valuable insights into the population's emotional state during public health crises.
• Findings underscore the adverse emotional impact of the pandemic and highlight the need for integrated mental health support.
• The study demonstrates the potential of social media data for informing public health policy and mental health infrastructure.