These are the SLO classifiers each with the supporting CSIRO-internal datasets.
Per an email from Brian Jin, Feb 6, the data used by the classifiers is collected by a data publisher that reads tweets from ESA (Emergency Situation Awareness) with the queries :
- adani : stopadani, adani, goadani
- bhp : bhpbilliton, bhp
- cuesta : cuestacoal, "cuesta coal", cuesta,cqc
- fortescue : fortescuenews, "fortescue metals", fortescue
- riotinto : riotinto, "rio tinto"
- newmontmining : newmont, "newmont mining"
- santos : santosltd, santos
- oilsearch : oilsearchltd, "oil search"
- woodside : woodside,woodsideenergy,"woodside petroleum","woodside energy"
- ilukaresources : iluka,ilukaresources, "iluka resources"
- whitehavencoal : whitehaven,whitehavencoal,"whitehaven coal"
If tweet text includes the given queries, the data publisher pushes the tweet into the data pipeline for the further processing.