To read this content please select one of the options below:

Measuring climate change on Twitter using Google’s algorithm: perception and events

Ahmed Abdeen Hamed (EPSCoR, University of Vermont, Burlington, Vermont, USA)
Alexa A. Ayer (Rubenstein School, University of Vermont, Burlington, Vermont, USA)
Eric M. Clark (Department of Computer Mathematics, University of Vermont, Burlington, Vermont, USA)
Erin A. Irons (EPSCoR, University of Vermont, Burlington, Vermont, USA)
Grant T. Taylor (EPSCoR, University of Vermont, Burlington, Vermont, USA)
Asim Zia (EPSCoR, University of Vermont, Burlington, Vermont, USA)

International Journal of Web Information Systems

ISSN: 1744-0084

Article publication date: 16 November 2015

772

Abstract

Purpose

The purpose of this paper is to test the hypothesis of whether more complex and emergent hashtags can be sufficient pointers to climate change events. Human-induced climate change is one of this century’s greatest unbalancing forces to have affected our planet. Capturing the public awareness of climate change on Twitter has proven to be significant. In a previous research, it was demonstrated by the authors that public awareness is prominently expressed in the form of hashtags that uses more than one bigram (i.e. a climate change term). The research finding showed that this awareness is expressed by more complex terms (e.g. “climate change”). It was learned that the awareness was dominantly expressed using the hashtag: #ClimateChange.

Design/methodology/approach

The methods demonstrated here use objective computational approaches [i.e. Google’s ranking algorithm and Information Retrieval measures (e.g. TFIDF)] to detect and rank the emerging events.

Findings

The results shows a clear significant evidence for the events signaled using emergent hashtags and how globally influential they are. The research detected the Earth Day, 2015, which was signaled using the hashtag #EarthDay. Clearly, this is a day that is globally observed by the worldwide population.

Originality/value

It was proven that these computational methods eliminate the subjectivity errors associated with humans and provide inexpensive solution for event detection on Twitter. Indeed, the approach used here can also be applicable to other types of event detections, beyond climate change, and surely applicable to other social media platforms that support the use of hashtags (e.g. Facebook). The paper explains, in great detail, the methods and all the numerous events detected.

Keywords

Acknowledgements

This research is partly funded by Vermont EPSCoR award number EPS-1101317. We acknowledge the Vermont Advanced Computing Core which is supported by NASA (NNX 06AC88G), at the University of Vermont for providing High Performance Computing resources that have contributed to the research results reported within this paper. We particularly thank Dr Peter Dodds and Dr Chris Danforth for the Twitter garden-hose historical dataset.

Citation

Hamed, A.A., Ayer, A.A., Clark, E.M., Irons, E.A., Taylor, G.T. and Zia, A. (2015), "Measuring climate change on Twitter using Google’s algorithm: perception and events", International Journal of Web Information Systems, Vol. 11 No. 4, pp. 527-544. https://doi.org/10.1108/IJWIS-08-2015-0025

Publisher

:

Emerald Group Publishing Limited

Copyright © 2015, Emerald Group Publishing Limited

Related articles