Note: Ideally we’d add an overview by reason (i.e plagiarism, low value, etc.), but since reasons are not typified, I figure the latter would not be feasible.
If the text of reports were made public, or if the text of reports was given to a select group of people to prepare a report, or if the text was tokenized and made public, it should not be difficult to classify each report.
If particular rules could be cited by moderators when taking moderation actions, this information could potentially be used to make improvements to the forum.