Passer au contenu principal
logo

2018 Conference

du 20 au 23 June 2018

Washington, DC

Please wait while schedule loads.

Machine Learning Techniques to Synthesize Multi-disciplinary Fields as Illustrated through a Meta-Analysis of Adaptive Capacity to Climate Change

samedi 23 juin 2018 à 09:00–10:30 EDT
SIS 333
Type of Session

Individual Paper Presentation

Abstract

Many of the world’s most pressing environmental issues require insights from multiple disciplines working together. Multi-disciplinary research provides new methods and insights, but it also raises the risk of fragmentation, and a fragmented field is less likely to provide coherent guidance to practitioners.  Synthesis and meta-analysis can help mitigate fragmentation, but meta-analysis in the social sciences can be difficult as results are often reported in narrative or qualitative format. The digital humanities, an interdisciplinary blend of computer science and humanities, offer several techniques – such as algorithmic text mining and natural language processing – to analyze patterns within narrative texts. By applying these techniques to the academic literature, or to qualitative data sources such as interview transcripts, documents, newspaper articles, or social media posts, researchers can explore patterns not visible to the manual reader and synthesize large textual data sets. These techniques are illustrated through application to climate adaptation research.

Adapting to global climate change and building community resilience to slow-onset change and extreme weather events will require people and groups with the capacity to change – an ability called adaptive capacity. This study uses machine learning techniques to analyze 88% of all English-language academic and practitioner publications on adaptive capacity, and to build a conceptual model of adaptive capacity. Results support the creation of a framework, the Adaptive Capacities Framework (ACF), that may be used to assess capacity and prioritize resource allocation. The ACF is novel for its emphasis on functional abilities rather than possession of assets, and for its ability to apply at multiple scales and thereby enable cross-scale comparisons and research on inter-scale interactions. A new theory, based on these results, is proposed to reconcile a theoretical divide within the field, raise new questions about theoretical limits of adaptability and the rate of social adaptation, and suggest an alternative approach for practitioners to allocate resources.

Primary Contact

A.R. Siders, Stanford University

Presenters

A.R. Siders, Stanford University

Co-Authors

Chair, Facilitator, Or Moderators

Discussants

Workshop Leaders

Chargement en cours …