What is known about climate change? A knowledge graph approach

Thesis

The latest findings from the Intergovernmental Panel on Climate Change Sixth Assessment Report retells a story already told in 1979 in greater detail --- that climate change is happening because of human causes. However, the progress on addressing the climate crisis hardly corresponds to the media's nearly 40 years of news coverage. A widely cited barrier to progress is the lack of knowledge to comprehend the various interacting aspects of climate change. As a dynamic complex system, climate change is a deeply relational systems-level problem that requires communication and mental models capable of communicating system complexities. To establish the extent of relations and knowledge formations by past communications, this paper offers a methodological innovation enabling network analysis of unstructured texts. This research developed a reproducible automated semantic network approach using natural language processing and open information extraction tools to construct knowledge graphs (KGs) on climate change news coverage in 5 countries using news articles $(N = 19,684)$ from 1990 to 2020. Country comparisons were carried out cross-sectionally and longitudinally. The findings show that the KGs are structurally similar while static but differed substantially in longitudinal patterns. Countries performing better in climate commitments saw a uniform increase in the salience and embedded of concepts, while in those less progressive countries, concepts become salient sporadically. The study concludes by recommending a concerted communication model that gives equal and parallel issue-attention to the causal, action, and physical characteristic and consequence frames of climate change to facilitate understanding of the system complexity of climate change.

Authors: Zhang, T

• DOI: 10.5287/ora-9eopbagnv
• Type of award: MSc
• Level of award: Masters
• Awarding institution: University of Oxford
• Language: English
• Keywords: frame analysis; semantic network; media framing; knowledge graphs; network analysis; open information extraction; climate change communication
• Subjects: Climate; Natural language processing (Computer science); Semantic networks (Information theory); Communication