EN VI RO N MEN TA L MONITORING USING
SOCIAL MEDIA THE GREAT BARRIER REEF is an iconic natural attraction that is visited by several million people every year. Many Reef visitors use social media to share their experiences and perceptions. This research examines whether online data can be used for environmental monitoring purposes.
Different types of analysis are applied to a series of social media data. They rely on domain knowledge to inform machine learning, deep learning, and neural networks to derive meaningful results.
Social media provide large volumes of information at no or little cost, but the data requires considerable cleaning and filtering to extract useful insights.
Twitter: >700 tweets globally mention the GBR every day.
Weibo: >50 posts mention the GBR per day.
Twitter: >1,200 tweets posted from the GBR region per day.
Facebook: >25 posts and responses per day across 13 commercial Fb pages.
IMAGE PROCESSING Flickr images: >50 images tagging GBR per day.
S ea T ur tle Parro t
SENTIMENT ANALYSIS It is possible to use social media data alongside more traditional forms of data collection.
Average Facebook sentiment
Average Twitter sentiment 0.7
COST OF COLLECTION
0.5 0.4 0.3 0.2 0.1
RY UA BR
R NO VE
R BE EM PT SE
EC (E TIV .G E . T SE W NS IT IN (E TE G .G H R) .E U YE MA O N N SE TH N E SO RE R EF S CI ) TI ZE N SC IE NC E PR O M FE ( S O N SS CI IT IO EN OR NA TI IN L ST G S)
Funded by: National Environment Science Program, Tropical Water Quality Hub. Prof Susanne Becken, Prof Bela Stantic, Prof Rod Connolly, Griffith University. For more information contact [email protected]