We would like to know how the confidence levels affect results. So we could classify to:
Fake news, True news, Can't give prediction
We would like to know how claim size affects the confidence. Possible measures:
posts in claim
authors in claim
edges in claim
Also we would like to know:
If we measure only claims above some confidence level - our results gets better?
We would like to know how the confidence levels affect results. So we could classify to: Fake news, True news, Can't give prediction We would like to know how claim size affects the confidence. Possible measures:
Also we would like to know: If we measure only claims above some confidence level - our results gets better?