基于数据挖掘的公众环境风险感知焦点分析OA

Ensuring a precise response to the public's environmental demands is a vital prerequisite for promoting the modernization of environmental governance.Based on ecological and environmental complaint data from Sichuan Province,a hybrid research framework of"qualitative coding and quantitative modeling"was constructed in this study.Methods including word frequency analysis,Latent Dirichlet Allocation(LDA)topic modeling,and sentiment analysis were synthesized to systematically analyze the structural features and evolution patterns of public environmental risk perception.It was found that public risk perception exhibited the core characteristics of"sensory dominance,government reliance,and surface-level reasoning".The dimensions of direct perception and environmental change account for the highest proportions,and a tendency among the public to attribute governance responsibility to the government was observed.Furthermore,the public opinion analysis revealed the patterns of pollution type coupling and temporal and spatial differentiation,presenting a"high in the east and low in the west,concentrated in the center"pattern,indicating that risk perception was significantly influenced by spatial proximity,temporal persistence,and the cumulative effect of multiple pollutants.This study provides a scientific basis for the government to formulate differentiated risk communication and precise governance strategies.