Air Pollution News Sentiment: Seasonal Trends' Relation with PM10 Levels

Abstract

Air pollution remains a major public health concern globally, especially in the Western Balkan countries facing severe challenges with the air quality problem. This study investigates the relationship between air quality, news media sentiment, and public discourse in Macedonia over a ten-year period (2014-2023). We employed sentiment analysis to examine the emotional tone of news coverage related to air pollution, and topic modeling to uncover recurring themes within news articles. Our analysis revealed a distinct seasonal pattern, with negative media sentiment peaking during winter months when PM10 levels were the highest. This finding aligns with the increased reliance on polluting fuels for winter heating. Interestingly, despite a stable number of neutral articles, a rise in positive sentiment articles suggests a potential decrease in pollution levels or the effectiveness of new government policies. We identified recurring topics like air quality concerns in specific cities, public unease regarding factories, and ongoing scrutiny of government policies. Emerging topics included the impact of the COVID19 pandemic on air quality, public discourse surrounding heating practices, and growing concerns about waste management. This study contributes to a deeper understanding of the complex interplay between air pollution data, public discourse, and media framing, offering valuable insights for policymakers and media outlets in Macedonia.