Spatio-temporal difference analysis in climate change topics and sentiment orientation: Based on LDA and BiLSTM model

Abstract

Sentiment orientation, as a psychological variable, has a significant effect on climate change issue communication. In this study, 169, 592 blog posts related to climate change on Weibo are crawled. Combining Bidirectional Long Short Term Memory (BiLSTM) text sentiment classification model and Latent Dirichlet Allocation (LDA) topic model, spatio-temporal difference analysis of the blog posts, public attention, hot topics and sentiment orientation to climate change are discussed. Empirical results show that: (1) Hot topics of climate change can be clustered into four categories. And the public pays more attention to Topic 1 (green development and energy transformation) and Topic 3 (the adverse effects of climate change on human life). (2) Most Chinese present positive sentiment toward climate change, which implies relatively high public support or satisfaction with the relevant issues. (3) The driving factors of the public's positive and negative sentiments about climate change are obtained.