Skip to main content

Table 2 Annotation classes

From: Deep learning for pollen allergy surveillance from twitter in Australia

Class

ID

Description

Example

Informative

1

Personal reporting including: symptoms (a), treatments (b) or both (c).

‘90% sure we have either developed hayfever. My eyes feel so dry and my nose is running whenever i go outside’ (a)

   

‘my hayfever better watch out today bc ate a spoonful of local honey AND took antihistamines’ (b)

   

‘Kings Cross triggered my hayfever today and now I can’t stop sneezing. I need a zyrtec’ (c)

 

2

General personal reporting.

‘Hayfever suuucks’

Non-Informative

3

Public broadcasting including: marketing (a), news (b), warnings (c) etc.

‘Himalayan Inhaler to help with asthma & hayfever #himalayansalt #onlineshopping’ (a)

   

‘Scientists in Melbourne have made a breakthrough that could change the lives of asthma and hay fever sufferers’ (b)

   

‘Be careful Victorians: thunderstorm asthma warning. Stay safe. #hayfever #health #bom #weather’ (c)

 

4

Unrelated/Ambiguous.

‘Very annoying. Could be a preservative or maybe hay fever.’