A valid probability model must:
account for all events;
assign probabilities for each event;
sum to one.
In our fake_news example, the prior probability model is defined as:
fake_news
P(is fake)=0.4 and P(is real)=0.6