Exploring public attention about green consumption on Sina Weibo: Using text mining and deep learning

Abstract

Achieving the goal of carbon neutrality and carbon peak as scheduled puts forward new demands for the green transition of low-carbon lifestyle in Chinese society. In-depth practice of green consumption (GC) behavior can effectively promote the supply-side and consumption-side emission reduction work, but the phenomenon of “high awareness, low practice” is widespread in GC. The causes of consumers' low practice of GC need to be analyzed from the perspective of time and space from the actual media data. Furthermore, this process assists policymakers and stakeholders to understand the general attitude of the public towards GC, clarifying the propagation path of public emotions and the source of negative emotions. Based on the data from Sina Weibo, this paper applied text mining, a hybrid model of convolutional neural network and long and short-term memory neural network to analyze the public's attention, sentiment tendency and hot topics on GC. The results show that the vast majority of the Chinese public has a positive attitude toward GC; women and economically developed regions are more concerned about GC; the drivers of positive public sentiment toward GC include environmental awareness education, air pollution prevention and control, and online shopping; high green product prices, excessive time costs, chaotic sharing economy and one-size-fits-all solutions lead to negative public sentiment toward GC. By providing public sentiment analysis of GC, this research would assist decision-makers to understand the dissemination mechanism of public will in social media and clarify targeted solutions, which is of great significance for policy formulation and improvement.