If i understood your question right, you're building a prediction model to anticipate the reaction of SE.AI to certain events. For example, if a user is posting a text about problems in the error-rate in the RNN network into the form, the following reaction can be there. First, the question is ignored ot a moderator can flag it as offtopic, the posting gets lots of upvotes or the posting gets an answer which brings the forum forward.
Creating such a prediction model for an online-discussion room is an important step because it helps to avoids pitfalls. If somebody is sure, that questions about Artificial Intelligence in movies (only an example) are off-topic on this website, he can adapt his behavior and won't post such questions. If the prediction model is right, the user can increase his score and avoid off-topic flags given by others. So the question is: what is the community doing in reaction to a post? are rules available for saying so in advance?
There are some measurements available. At first, every online-forum has an archive which allows to understand reactions in the past, secondly every forum has a certain form of users interacting with the forum. In the case of SE.AI most users are interested in new technology which is Artificial Intelligence and they come with a background in computer science. But, other users are not focused on academic AI, instead they have listen to the AI Influencer Sirai Raval, who is marketing neural networks to everybody.