How to label data for model training?

Question

Machine learning

Good day , I am
faced with the task of classifying messages from social networks into suicidal and non-suicidal
, and I don’t really understand how to do this, it
seems to be the task of text classification, about which there is a lot of information on the Internet,
but on the other hand, wherever you look, they all learn from ready-made data,
and I have a problem precisely in these ready-made data
, in my database I have downloaded a lot of messages from the wall from VK with their pictures,
but now how to mark it all up?
bypass manually even 1000 messages? or are there better ways? what criteria to set for their classification?

Answer 1

People mark up, either on their own or for their money.
Here you can throw it on Toloka . But IMHO they pay such pennies there that you can wait six months for the result.