Nonlinear impact of digital economy on carbon intensity: the moderating role of low-carbon regulation

The development of the digital economy is an effective way to mitigate the carbon emission problem in the broader setting of the significant data era and green development. Based on the panel data of 271 cities in China from 2011 to 2019, this paper constructs a bidirectional fixed model to analyze the nonlinear effect of the digital economy (DE) on carbon intensity (CI) and the moderating role of low-carbon regulation from theoretical and empirical perspectives. The results show that (1) DE has an enormous inverted U-shaped impact on CI. The findings remain after introducing instrumental variables to mitigate endogeneity and robustness tests. (2) Low-carbon regulation (CP) can strengthen the inverted U-shaped impact between the two and shift the inflection point to the left. (3) Heterogeneity analysis shows that the inverted U-shaped effect of DE on CI is more significant in the central and western regions, high human capital (HC) regions, and high urbanization regions. (4) The mediating effect of energy mix (EM) and green technology innovation (GTI) still hold after introducing instrumental variables to alleviate the endogenous effect of the intermediary effect. This study suggests that the adoption of carbon emission reduction strategies, which will more effectively lower carbon intensity CI, should go hand in hand with the development of DE.

Attention, sentiments and emotions towards emerging climate technologies on Twitter

Public perception of emerging climate technologies, such as greenhouse gas removal (GGR) and solar radiation management (SRM), will strongly influence their future development and deployment. Studying perceptions of these technologies with traditional survey methods is challenging, because they are largely unknown to the public. Social media data provides a complementary line of evidence by allowing for retrospective analysis of how individuals share their unsolicited opinions. Our large-scale, comparative study of 1.5 million tweets covers 16 GGR and SRM technologies and uses state-of-the-art deep learning models to show how attention, and expressions of sentiment and emotion developed between 2006 and 2021. We find that in recent years, attention has shifted from general geoengineering themes to specific GGR methods. On the other hand, there is little attention to specific SRM technologies and they often coincide with conspiracy narratives. Sentiments and emotions in GGR tweets tend to be more positive, particularly for methods perceived to be natural, but are more negative when framed in the geoengineering context.

Topic sentiment analysis based on deep neural network using document embedding technique

Sentiment Analysis (SA) is a domain- or topic-dependent task since polarity terms convey different sentiments in various domains. Hence, machine learning models trained on a specific domain cannot be employed in other domains, and existing domain-independent lexicons cannot correctly recognize the polarity of domain-specific polarity terms. Conventional approaches of Topic Sentiment Analysis perform Topic Modeling (TM) and SA sequentially, utilizing the previously trained models on irrelevant datasets for classifying sentiments that cannot provide acceptable accuracy. However, some researchers perform TM and SA simultaneously using topic-sentiment joint models, which require a list of seeds and their sentiments from widely used domain-independent lexicons. As a result, these methods cannot find the polarity of domain-specific terms correctly. This paper proposes a novel supervised hybrid TSA approach, called Embedding Topic Sentiment Analysis using Deep Neural Networks (ETSANet), that extracts the semantic relationships between the hidden topics and the training dataset using Semantically Topic-Related Documents Finder (STRDF). STRDF discovers those training documents in the same context as the topic based on the semantic relationships between the Semantic Topic Vector, a newly introduced concept that encompasses the semantic aspects of a topic, and the training dataset. Then, a hybrid CNN-GRU model is trained by these semantically topic-related documents. Moreover, a hybrid metaheuristic method utilizing Grey Wolf Optimization and Whale Optimization Algorithm is employed to fine-tune the hyperparameters of the CNN-GRU network. The evaluation results demonstrate that ETSANet increases the accuracy of the state-of-the-art methods by 1.92%.

Attitudes towards Global Warming on Twitter

Public opinion surveys over the past thirty years show that public opinion is split on the issue of global warming. One of the problems with “solicited” opinion polls is that the findings may be selectively interpreted in favour of the political goals of a particular interest group. To gain a better understanding of the general public’s unsolicited responses to climate change news, the current study examined Twitter messages containing the words “global warming” spanning 16 months. Using a framework combining a sentiment analysis technique, Hedonometer from the perspective of natural language processing, and Appraisal Theory from a discourse analysis perspective, the study shows that the demonstrated happiness level in tweets containing the words “global warming” is consistently lower than the general level on Twitter, due to increased use of negative words and decreased use of positive words. The Appraisal analysis shows that Appreciation is used most frequently and Affect least.