Portal de Periódicos da CAPES

Leveraging Large Language Models for Analyzing Climate Change Mitigation Technology Dissemination: A Case Study of Wind Power in the UK

2024; IOS Press; Linguagem: Inglês

10.3233/atde240955

2352-7528

Kenji Yamada, Kosaku Nakano, Rokuro Tomita, Hiroyoshi Iwata, Seita Emori, Kenji Tanaka,

Computational and Text Analysis Methods

The progress of global warming requires countries around the world to accelerate measures to combat climate change. Under these circumstances, the adoption and dissemination of climate change mitigation technologies have become urgent tasks. However, these processes always face numerous challenges, such as high costs and resistance from local communities. Moreover, understanding the status of technology implementation, including these challenges, requires considerable time and effort. This study addresses this issue by applying Large Language Models (LLMs) to identify the key issues related to climate change mitigation technologies. We propose two methodologies: the “document-based approach,” which consists of document summarization, clustering, and topic labeling, and the “topic-based approach,” which consists of topic extraction and counting by classification. In this study, we apply these methodologies to wind power generation using two news media data and analyze the topic trends. There are three main findings: 1) different patterns of topic emergence, 2) detailed trends in discussions about the environmental impacts of wind power, and 3) different characteristics between the two media sources. In addition to these three findings, one of the values of this study is to provide an example of the use of LLMs to monitor the adoption and dissemination of climate change mitigation technologies.