MathOverflow tags + others ?

Question

I have re-tagged two questions today, 0.999=1? and the issue of the divergence of the harmonic series as "calculus" because I don't think they are real-analysis questions. However, when a real-analysis question comes up, I think it should be tagged in a consistent manner with MO for the sake of continuity and for the sake of those persons here pursuing a career in research-level mathematics. MO is a very successful site and we should seriously consider adopting what works and what is industry standard (as has been pointed out by Noah Snyder in reference to the arxive standard).

Can we use MO tags for questions more typical of advanced undergrad/graduate coursework and make up broader or more vague tags for questions typical of lower-level undergrad coursework?

Answer 1

I'm fairly certain that many of the arXiv tags won't be used much (e.g. [ag.algebraic-geometry]). I also think it would be better simply to use [algebraic-geometry] than [ag.algebraic-geometry], since we are not focusing on research here, and matching the arXiv is less of a concern.

I support [big-list] and [soft-question]. Some people dislike those sorts of questions; they can choose to ignore them as a result. This may reduce friction on math.SE.

The use of a [high-school] tag may be appropriate when the question is asked directly motivated by material from a high school class. For instance, anything up to elementary calculus questions can be tagged here. Nevertheless, there are sophisticated questions that use only high school mathematics in novel ways, such as in hard olympiad problems; I would consider this tag inappropriate in that case. Some of those questions may even be MO-level! So the sophistication of the answer should be considered as well.

Answer 2

I think whoever moderates will end up needing to re-tag a lot of things at the beginning, based on general guidelines reached by consensus. Eventually what tagging systems work will become clearer, and that new knowledge should be codified in the FAQ. (Those who don't read the FAQ, and who haven't used MO or similar sites, will often end up tagging in an arbitrary and subjective manner that does not reflect how regular users of the site have come to expect it to work, and their tags will thus have to be redone often.)

Answer 3

I think the [soft-question] and [big-list] tags are excellent, and I am happy to see the effort Harry has gone to in bringing that across.

I doubt the ArXiv-based two-letter prefix tags make sense here: it only half works on MO. ArXiv tags make sense when there is a community of researchers who might be interested in some issue: most of the questions here are matters that any graduates worth their salt will be able to figure out how to answer.

I retagged the Does .99999...=1? question to [decimal] [number]: limits might be part of the right answer, but they are not part of the question.

Answer 4

I am in support of using MO's arXiv-based tags for questions at the advanced undergraduate-early graduate level. The SE platform allows users to filter what they see on the front page by ignoring tags they're not interested in, and I think this will be an absolutely crucial method for increasing satisfaction with the site across multiple types of users who are looking for rather different things.

A big question I see here is how to label questions that are more basic than something that can be given an arXiv classification. There's a basic question on the main page right now about a dice game labeled - inappropriately, I believe - "pr.probability-theory," and I think being loose with those classifications will frustrate users.

The thing about tags based on a highly subjective judgment of level ("high-school," "basic," etc.) is that 1) These tags will most likely have to be adjusted, since the asker will often tag inappropriately, and 2) We've already had accusations of elitism flying around, and however well-intentioned and non-judgmental contributors try to be, there will be those who bristle at level tags, especially when they are necessarily imposed externally in many cases.

Edit: I agree that the use of the two-letter prefix is unnecessary here, and will complicate things for non-arXiv users who still want to ask undergrad-etc. level questions. Some of the arXiv tags won't get used much; perhaps tags should correspond to standard names of subjects as labeled in course names. For example, real-analysis is, I think, a reasonable label, despite not being an arXiv classification explicitly.

I would like, however, to try to make a system where, for example, standard questions about the limit of a sequence or sum of a particular series are not labeled real-analysis: calculus would be appropriate, I think.

As another example, we're going to get a lot of computational matrix algebra problems, so I might advocate that questions about abstract vector spaces be labeled abstract-algebra and not linear-algebra.

Answer 5

This sounds good to me, but not everyone on this site is familiar with the MathOverflow tagging scheme. Is there an tagging index/FAQ on MO's meta site or something that could be reproduced here for reference?

Also, no reason to limit the MO tags to advanced questions. If there's a basic number theory question, I think it would be most sensible to use the same number theory tag as any other question; just tack on an elementary tag.

Answer 6

I'm not familiar with MO's tags.

Don't know if it's possible, but can we "import" all the tags from MO to here so they appear automatically when trying to come up with a tag? That would solve a lot of the early problems. I don't think a FAQ entry will help much, since most people won't read it anyway.