National Center for High-Performance Computing, Hsinchu City 300, Taiwan.
Program of Technology Management, Department of Industrial Education, National Taiwan Normal University, Taipei 106, Taiwan.
Int J Environ Res Public Health. 2021 May 15;18(10):5270. doi: 10.3390/ijerph18105270.
With growing scientific evidence showing the harmful impact of air pollution on the environment and individuals' health in modern societies, public concern about air pollution has become a central focus of the development of air pollution prevention policy. Past research has shown that social media is a useful tool for collecting data about public opinion and conducting analysis of air pollution. In contrast to statistical sampling based on survey approaches, data retrieved from social media can provide direct information about behavior and capture long-term data being generated by the public. However, there is a lack of studies on how to mine social media to gain valuable insights into the public's pro-environmental behavior. Therefore, research is needed to integrate information retrieved from social media sites into an established theoretical framework on environmental behaviors. Thus, the aim of this paper is to construct a theoretical model by integrating social media mining into a value-belief-norm model of public concerns about air pollution. We propose a hybrid method that integrates text mining, topic modeling, hierarchical cluster analysis, and partial least squares structural equation modelling (PLS-SEM). We retrieved data regarding public concerns about air pollution from social media sites. We classified the topics using hierarchical cluster analysis and interpreted the results in terms of the value-belief-norm theoretical framework, which encompasses egoistic concerns, altruistic concerns, biospheric concerns, and adaptation strategies regarding air pollution. Then, we used PLS-SEM to confirm the causal relationships and the effects of mediation. An empirical study based on the concerns of Taiwanese social media users about air pollution was used to demonstrate the feasibility of the proposed framework in general and to examine gender differences in particular. Based on the results of the empirical studies, we confirmed the robust effects of egoistic, altruistic, and biospheric concerns of public impact on adaptation strategies. Additionally, we found that gender differences can moderate the causal relationship between egoistic concerns, altruistic concerns, and adaptation strategies. These results demonstrate the effectiveness of enhancing perceptions of air pollution and environmental sustainability by the public. The results of the analysis can serve as a basis for environmental policy and environmental education strategies.
随着越来越多的科学证据表明，空气污染对现代社会环境和个人健康的有害影响，公众对空气污染的关注已成为空气污染防治政策发展的核心焦点。过去的研究表明，社交媒体是收集公众意见数据和分析空气污染的有用工具。与基于调查方法的统计抽样相比，从社交媒体中检索的数据可以提供有关行为的直接信息，并捕获公众生成的长期数据。然而，关于如何从社交媒体中挖掘有价值的见解来了解公众的环保行为，研究还很缺乏。因此，需要将从社交媒体网站中检索到的信息纳入关于环境行为的既定理论框架中进行研究。因此，本文的目的是通过将社交媒体挖掘纳入公众对空气污染的关注的价值-信仰-规范模型来构建一个理论模型。我们提出了一种混合方法，该方法将文本挖掘、主题建模、层次聚类分析和偏最小二乘结构方程模型（PLS-SEM）相结合。我们从社交媒体网站中检索了有关公众对空气污染的关注的数据。我们使用层次聚类分析对主题进行分类，并根据价值-信仰-规范理论框架来解释结果，该框架包括自我关注、利他关注、生物关注和针对空气污染的适应策略。然后，我们使用 PLS-SEM 来确认因果关系和中介的影响。基于台湾社交媒体用户对空气污染的关注的实证研究，以一般的方式展示了所提出框架的可行性，并特别考察了性别差异。根据实证研究的结果，我们确认了公众对空气污染的自我关注、利他关注和生物关注对适应策略的有力影响。此外，我们发现性别差异可以调节自我关注、利他关注和适应策略之间的因果关系。这些结果表明，通过增强公众对空气污染和环境可持续性的认识，可以提高其效果。分析的结果可以作为环境政策和环境教育策略的基础。
