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

Human induced climate change is one of this century's greatest unbalancing forces that affect our planet. Capturing the public awareness of climate change on Twitter has proven to be significant. We demonstrated in our previous research that public awareness is prominently expressed in the form of hashtags that use 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??). We learned that the awareness was dominantly expressed using the hashtags: (#ClimateChange) respectively. In this paper, we test the hypothesis of whether more complex and emergent hashtags can be sufficient pointers to climate change events.

The methods we demonstrate 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.

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

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