Unraveling environmental risk awareness in China over four decades from mass media and academic archives

Abstract Environmental risks have become significant impediments to economic and social development. Increasing awareness of environmental risks can promote pro-environmental behaviors, thereby helping mitigate such risks. Currently, studies focusing on the extraction of environmental risk awareness are limited by several deficiencies, including the use of small data sets, brief analysis periods, and narrow research areas. To address these limitations, this study introduces a novel methodological framework that integrates Natural Language Processing (NLP) techniques with mass media and academic archives to extract environmental risk awareness over extended periods and large spatial scales. By applying geographic entity and focus extraction, coupled with environmental risk quantification, we examined nearly four decades of environmental risk awareness in China. The results reveal heightened awareness of natural disasters and climate change, while understanding of biological invasions and anthropogenic-related risks remains limited during the past four decades. Moreover, temporal, spatial, and thematic variations in the public’s and academics’ awareness are observed, with academia focusing more on systemic hazards and displaying a broader spatial distribution of awareness. National administrative centers significantly influence environmental risk awareness compared to provincial centers. Our research offers a useful tool for environmental risk awareness extraction, providing valuable insights for policy formulation and environmental conservation.