基于环境舆情文本挖掘的大气环境风险感知评价  

RISK PERCEPTION TOWARDS ATMOSPHERIC ENVIRONMENT BASED ON TEXT DATA MINING OF PUBLIC OPINIONS

作　　者：赵锐 刘婕 周宇羽 杨林川 ZHAO Rui;LIU Jie;ZHOU Yuyu;YANG Linchuan(Faculty of Geosciences and Environmental Engineering,Southwest Jiaotong University,Chengdu 611756,China;School of Architecture and Design,Southwest Jiaotong University,Chengdu 611756,China)

机构地区：[1]西南交通大学地球科学与环境工程学院,成都611756 [2]西南交通大学建筑与设计学院,成都611756

出　　处：《环境工程》2022年第4期209-216,共8页Environmental Engineering

基　　金：国家自然科学基金项目(41571520);四川省杰出青年基金项目(2019JDJQ0020);四川循环经济研究中心课题(XHJJ-2002,XHJJ-2005);成都市软科学研究项目(2020-RK00-00240-ZF,2020-RK00-00246-ZF);中央高校基本科研业务费专项资金(2682021ZTPY088)。

摘　　要：基于2015—2019年环境舆情信息,采用大数据挖掘与分析技术识别出公众对大气环境质量风险感知影响因素,建立了风险感知指标评价体系,分析了风险感知水平的时空变化特征。结果表明:1)影响公众风险感知的首要因素是个体健康担忧,其次是环境质量变化。2)公众风险感知水平具有显著的时空差异性;从时间上看,风险感知水平整体呈现波动上升趋势;从空间上看,风险感知水平呈现北方高南方低、东部高西部低的特征。3)河南、四川两省的风险感知水平在研究时间段内最高,且风险感知具有空间溢出效应。According to the public opinions on air quality during the period of 2015—2019, big data mining and analysis techniques were used to identify the factors influencing the public’s risk perception, establish a risk perception indicator system, and investigate the temporal and spatial characteristics regarding risk perception level. The results showed that: 1) The primary factor affecting public risk perception was the worry of individual health impact, followed by the impact of environmental quality change. 2) The level of risk perception had significant temporal and spatial differences. It fluctuated and raised in the defined time period. From a spatial perspective, the level of risk perception was high in the north and low in the south, while high in the east and low in the west. 3) The risk perception level in Henan Province and Sichuan Province was the highest in the defined time period, and the risk perception had apparently spatial spillover effect.

关 键 词：环境舆情 大数据 数据挖掘 风险感知 时空分布 

分 类 号：X823[环境科学与工程—环境工程]