Public Perception of Climate Change in Alaska: A Case Study of Opinion-Mining using Twitter. GI_Forum|GI_Forum 2018, Volume 1 |

The Arctic, and with it the State of Alaska, USA, is an area highly impacted by climate change. Changing environmental conditions have started to impact local communities, causing a need for changes ranging from new infrastructure to the relocation of entire towns. These changes connected to rising...

Full description

Bibliographic Details
Published in:GI_Forum
Main Authors: Ristea, Alina, Bartsch, Annett, Resch, Bernd, Bergstedt, Helena
Format: Journal/Newspaper
Language:German
Published: oeaw 2018
Subjects:
Online Access:http://epub.oeaw.ac.at/?arp=0x00390cc2
https://doi.org/10.1553/giscience2018_01_s47
Description
Summary:The Arctic, and with it the State of Alaska, USA, is an area highly impacted by climate change. Changing environmental conditions have started to impact local communities, causing a need for changes ranging from new infrastructure to the relocation of entire towns. These changes connected to rising temperatures have been shown to affect people’s overall health, and their mental health in particular. Previous studies using opinion-mining and Twitter data have focused on large areas, not distinguishing between regions within countries. In the course of the research presented in this paper, we analysed Twitter data for the period 2013–2017, from which we extracted opinions concerning climate change topics by applying sentiment analysis (polarity and feelings) and climate change dictionaries, on a 10 x 10 km grid for the State of Alaska, USA. The number of climate change-relevant tweets was found to be much lower than reported in previous studies, where the USA was only considered in its entirety. After applying a topic-modelling approach, we found little difference between the spatial distributions of hotspots for the different climate change topics. A comparison with population data showed considerable biases towards English-speaking communities, tweets in indigenous languages being excluded when pre-defined dictionaries in English were used.