I read that Jira Service Desk uses machine learning to optimise the KB articles which are being presented.
Is this still true? I have read conflicting information suggesting it is purely based on what is in the summary field.
