Figure 1 . Conceptual overview of how the topic modeling analysis feeds into the structured futuring process, including worldbuilding and story creation
Figure 2. Results of the latent Dirichlet allocation, with thematic dominance and clustering visualized on the left, and the baseline frequency of keywords across the entire corpus depicted on the right. This was generated using the freely available pyLDAvis software in the GENSIM Python package.
Figure 3. Summary of the story-based scenario creation procedure for Topic #5 ‘Concession 60’. Steps begin at the top left, and can be read left-to-right, and top-to-bottom.
Figure 4. Examples of visual art for two of the story-based scenarios. The top panel is an image inspired by ‘Concession 60’ (i.e., Topic #5), and the bottom panel is an image inspired by ‘Assisted Migration’ (i.e., Topic #9). Artwork is used with permission by Patrick W. Keys and Fabio Comin. All rights reserved.
Table 1 . Summary of topic clusters with corresponding top 30 keywords in that cluster with terms listed in descending order of relevance within the topic (i.e., the first term is the most salient to the topic).
Table 2 . Overview of stories, including theme, title, and brief description of the world and story.
Table 3 . Comparison of a sample of existing Arctic scenario narratives with the ten scenarios produced in this analysis