7
$\begingroup$

I've noticed several questions popping up recently on the topic of Bayesian Optimization. Many of the questions are bizarre, full of misconceptions, and (as some moderators have noted) perfect examples of XY problems. But these questions aren't always closeable per se. If we dealt with just one such instance, an acceptable answer would be to gently correct their understanding. But we sometimes deal with 100s of undertrained statisticians participating in a free, possibly low quality MOOC. If we have 1000s of questions from them it doesn't make sense for the site to contain 1000s of copies of the same general answer. "X is wrong, thus Y doesn't matter".

This leads me to ask a two-part question (knowingly violating the site recommended rules);

  • Is there a way to detect trends in questions when a "hot" topic surfaces, i.e. who are the askers, who are the answers, what is the view and vote activity on a particular topic, etc.?
  • how can we speed up review and possible closure knowing that hot topics tend to attract bad questions, and (as an olive branch) point these askers to higher quality sources on a particular subject?
$\endgroup$
1
  • 5
    $\begingroup$ I've noticed the flurry of BO questions, too. And also a sudden interest in questions related to sports gambling, probably because that's becoming more popular in the US. I think the best thing to do is identify and/or create a canonical Q&A that addresses the core concern. The trick is to think through what the canonical question should be, so that we can close the new arrivals as a duplicate. $\endgroup$
    – Sycorax Mod
    Jan 10 at 18:03

1 Answer 1

3
$\begingroup$

That's a valid apprehension. I would be extra vigilant to discern any patterns in the posts of any of my followed tags.

Is there a way to detect trends in questions when a "hot" topic surfaces, i.e., who are the askers, who are the answers, what is the view and vote activity on a particular topic, etc.?

You can check the popular tags to see which ones are more frequently used; then you can check for an individual tag its top users which would show you number of questions it received over the last 7 days, last 30 days and what percentage of those were unanswered. You would also be able to check the top askers and answerers both over the past one month and all time.

how can we speed up review and possible closure knowing that hot topics tend to attract bad questions, and (as an olive branch) point these askers to higher quality sources on a particular subject?

I would like you to check this related post What to do about "wrong" questions? in which the consensus was to

write an answer that essentially explains why the question itself is mistaken, even if such an answer does not answer the question

but also as whuber noted:

[...] find, edit, or create a canonical answer that addresses the problematic procedure you are concerned about and propose it as the duplicate.

I don't think we need an additional closure reason. Irrespective of how many such questions come, the effective way, in my opinion, is to, if it is an XY problem, make OP aware of the same and point them out a canonical post if this is available.

It might seem imperative that a "hot" topic would prompt new users ask such questions so the reviewers should be rather more attentive while reviewing first questions: implicit is the perennial request for more users from the community to come forward for volunteering in the review queues.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .