# Tag management 2021

New year, new tag management thread.

Rules of the game are basically the same:

• A particularly bad tag (a rule of thumb: “if I can't imagine a person classifying a tag as either interesting or ignored, I'm getting rid of it”),
• A tag that should be a synonym of an existing one,
• A tag that used for two or more completely unrelated things,
• A need to create a new tag.
• Wait a couple of days before implementing a suggestion.
• After the problem described in an answer is resolved, please edit it to say so.

Also, note that one may use [tag:calculus] for , i.e. tags on the main site, and [meta-tag:discussion] for , i.e. for tags on the meta site.

Note that, in some cases, it might be better to have a separate question. Typically this happens when a longer discussion is needed and several possible answers are expected, since answers to a question provide more space for a more detailed discussion than comments under an answer in this thread.

Resolved: The renaming and synonymizing has been done.

Proposal:

1. Rename the to .
2. Create the tag synonym $$\to$$ .
3. Create the tag synonym $$\to$$ .
4. Create the tag synonym $$\to$$ .

This is related to, but distinct from, J. W. Tanner's proposal, in that it does not create any new tags. The idea has been discussed in the tagging chatroom.

Unless there are any objections, I will implement this proposal in a week.

• I see that you've linked to chat - here is a link which shows the conversation that spans two days. (Although in the first day there are mainly examples of questions about dot product and your messages about updating the tag-info.) And, of course, if the discussion in chat continuous, we can add more to the same bookmark. Jan 7 at 9:36

We have around 3k questions associated with stars and bars, meaning that manual editing of these questions to include the tag is out of the question, but nearly every two days, someone asks a duplicate stars and bars problem, so perhaps having a tag for the future may not be a bad thing...

Resolved: Both tags renamed.

Proposal: Rename to , to

Nothing too crazy here, just some quick grammar fixes.

To ensure singular-versus-plural consistency between "mother" tags" and their "children",

Moreover, given , , , I propose that .

Proposal: Synonymize the meta tag with .

The main tag is synonymized with : this was raised in What is the usefulness of having "proof-verification" and "solution-verification" as different tags?.

I propose that we do the same to the corresponding meta tags. Currently, there are 8 meta questions tagged and 55 meta questions tagged , and 2 questions among these use both tags.

• I have suggested a synonym - so now users with sufficient score in the tag can vote for/against the synonym. (Of course, moderators can approve a synonym unilaterally, if they see that as the right course of action.) Oct 24 at 7:48

Proposal: create qr-decomposition

There are already lu-decomposition, cholesky-decomposition, schur-decomposition and other related tags on eigenvalues or the SVD. But not qr-decomposition, while there seem to be many questions on this.

Alternately, as suggested by Cameron Williams in this (now deleted) question, we might merge all those tags with the tag matrix-decomposition, which already exists. Not my preferred choice, but it's debatable.

• There are over 2K questions with the tag matrix-decomposition. In my opinion, the whole point of having tags of the form xyz-decomposition is to allow users to find the needles in the haystack more easily and more quickly. The cost of having a few extra tags should be negligible when compared to the cost of having members of Math SE wasting time needlessly — say, trying to find duplicates, failing to do so and answering questions that have already been answered dozens of times. Mar 9 at 1:35
• @RodrigodeAzevedo Agreed. As I said, not my preferred choice, but since someone suggested this in a question I deleted, I felt I had to mention this. Mar 9 at 8:08

At the moment, tag is a synonym for tag . Hence, I propose that

Say, on a question on the number of real roots of a given cubic polynomial, I would rather see tag than tag . The roots may be found by solving a cubic equation, but the question is on the polynomial itself, isn't it?

• . There are tags for group, sheaf, De Rham, etale, galois, local and equivariant cohomology, but no tag for Hochschild cohomology despite there being a decent number of questions about it. Hochschild cohomology is also highly relevant in the deformation theory of associative algebras, so there will likely be questions with the tag that would benefit from the inclusion of a tag.

• for questions relating to $$A_\infty$$, $$E_\infty$$ algebras etc. Or possibly just a single tag for $$A_\infty$$ algebras since I think this type of infinity algebra has the most questions related to it, and currently it seems like the most appropriate tags for differentiating it from questions about other types of algebras is to include the or perhaps tags, but it's entirely possible to discuss the technical aspects of $$A_\infty$$ algebras without ever mentioning either of these two things, even though they are conceptually relevant. So maybe something like like there is on mathoverflow.

• I saw you have created hochschild-cohomology, but I wonder if hochschild-homology-cohomology would be better. Aug 21 at 4:16
• @ArcticChar I did consider that, but the reason I didn't go with it is because the other tags about different kinds of (co)homology are only written as "such-and-such-cohomology" and so I figured it would be best to remain consistent with the other tags. Aug 22 at 15:29
• Having said that, maybe you're right since Hochschild homology seems to be much more prevalent, while things like group homology or Lie algebra homology seem to be less so, which is maybe why whoever made those tags decided to just write them as cohomology. Aug 22 at 15:34

I just noticed that in early August somebody created the tag . A list of the questions carrying the tag.

There's no tag wiki, hardly a surprise.

I suggest that we unceremoniously delete the tag. I don't think it helps much. The tag serves these questions well, possibly accompanied by or .

There are tags like , , for more limited scopes (I have some reservations about those as well, less so about the last item) and a more general .

It may be that my antipathy is directed more towards the practice of creating tags simply to create tags. There are many too specialized ones.

• I have added some SEDE queries in the Tagging chatroom. The tag was created and removed also in 2019, I found only a single questions from that year. Sep 8 at 5:25
• Thanks @MartinSleziak. Looks like I should look at that chatroom rather than these threads for more precise information :-) Sep 8 at 5:32

$$\textbf{Resolution}:\;$$ The tag dot-product was created by Xander Henderson;

$$\textbf{Proposal}:\;$$ Create the tag dot-product.

There is a tag for cross-product, but not one for dot-product.

There is a tag for inner-product-space, but that is more abstract than the usual dot product for $$\mathbb R^n$$.

• Seems like a synonym would be the ideal solution here, redirecting dot-product to inner-product-space But on the other hand this may confuse users who would be the target audience of dot-product
– Alexander Gruber Mod
Jan 5 at 1:29
• Are there really enough (unique) good questions about the dot product to justify a tag for it?
– Alexander Gruber Mod
Jan 5 at 1:30
• @AlexanderGruber I have added a few random examples in the tagging chatroom. Maybe somebody is able to find more of them - and looking at some of those question might help in deciding whether a separate tag for dot product would be suitable. Jan 5 at 6:42
• @AlexanderGruber I agree that a dot-product -> inner-product-space synonym would be ideal. I have edited the inner product space tag wiki in anticipation of such a move (i.e. I have highlighted the dot product a bit more). The goal of the edits was to prevent confusion on the part of "dot product" users.
– Xander Henderson Mod
Jan 5 at 13:21
• @FearfulSymmetry: I edited accordingly Jan 12 at 18:26
• @MartinSleziak: thanks for the good examples Jan 12 at 18:27

The tag has the following usage guidance:

Wiki: Hadamard product of $$A$$ and $$B$$, $$m\times n$$ matrices, Hadamard product is defined entrywise.
Excerpt: For questions about Hadamard product between two matrices, or it can concern analytic functions.

I feel that there should ideally be two different tags, one for the Hadamard product of matrices, and one for the Hadamard product of analytic functions. So, I propose the following:

Proposal:

1. Rename the current tag as .
2. Create a new tag .
3. Since a large majority of questions with the tag are about the matrix product (as far as I can tell), retag the few questions about products of analytic functions with the new tag proposed above.
• I brought this up around a couple of weeks ago in the Tagging chatroom. Though one of the messages was pinned on the starboard, it did not generate any discussion. Oct 2 at 12:45
• Personally I don't feel this is necessary, as the mechanics of the two concepts are similar enough and it doesn't seem there are a huge number of questions about this topic to warrant a second tag. Oct 12 at 16:25
• @Integrand I can understand that point of view. I think that if, at the conclusion of the discussion, it is decided that the tag shouldn't be split up, then at the very least the tag's usage guidance should be improved. Oct 12 at 16:47

Proposal: Add the meta tag deleted-comments.

This tag is present on meta.SE: https://meta.stackexchange.com/questions/tagged/deleted-comments.

The current meta tag includes

• deleted questions

As a result, to search for meta questions about deleted comments, one has to use the query string "deleted comments" under the tag . Such user experience can be improved by the creation of the meta tag deleted-comments.

• Among the currently existing tags, the question related to comments deletion might be tagged deletion+comment. Feb 5 at 21:30

How about a tag for L'Hopital's rule?

I have asked a few questions myself involving L Hopital's rule specifically, such as:

'Proof' that $f''(x)=\frac{f'(x)}{x}$

Does L'Hopital's rule imply that $\lim_{x\to a}\frac{f'(x)}{g'(x)} = \lim_{x\to a}\frac{f(x)}{g(x)}$ always?

It would have been useful to have a tag to show that L' hopital's rule was the crux of my questions.

I have also seen many questions that ask about how L Hopital's rule works in a specific case; again the main topic related to their questions is L' hopital's rule, and a L'Hopital's rule tag would have been useful to them. Search "l'hopital" on the search option and many such questions come up.

• Can you explain why you think that this tag is necessary? L'Hospital's rule is a tool for computing limits. Generally, someone here posts a question about a limit, and expects folk to find that limit for them. The tools used to do this are typically irrelevant. There are folk who want to do things without using certain tools, for which the limits-without-lhopital tag exists. I don't, however, see a compelling reason to include a "lohpital" tag. That said, my mind is open, and I'm willing to be convinced.
– Xander Henderson Mod
Jan 8 at 0:06
• @XanderHenderson please see my edit. Unfortuantely I do not have any links of other's questions that include L Hopital's rule at the moment. Jan 8 at 9:12
• So the clarification you have added now can be summarized as: The tag would be for questions about L'Hospital's rule - and not for computations of limits L'Hospital's rule. (Unless they are in some way specific to LH - for example, asking why LH works/does not work in some limit.) Jan 8 at 9:13
• I have collected some possible examples of such questions here in chat. Maybe also some other users might have suggestions (or at least say whether this would be the kind of questions suitable for that tag). Jan 8 at 9:54

Could the relation-algebra tag be reconsidered? It has a description suggesting that would suit relational-algebra. Two separate tags would seem appropriate.

I see this has been discussed years ago: Relation-algebra tag contains also questions about relational algebra

Wikipedia is very clear on the distinction between "relation algebra" and "relational algebra". While there is some overlap (as earlier discussion to above cited question shows) it is a distinction that is useful. References to sources on these two topics do show objectively that overall there is a difference in usage.

The fact that on MathOverflow the same tag "relation-algebra" is given the accepted definition, while here it is described as if it meant "relational-algebra" seems inconsistent and looks like evidence something is not quite right.

In a sense, rank-$$1$$ matrices are the "atoms" with which "composite" matrices are built. Moreover, factoring polynomials and solving some combinatorial optimization problems can be reduced to finding a rank-$$1$$ matrix. Thus, I propose that tag be created.

• Perhaps it is worth mentioning that this tag was created and removed in December 2020. I have asked about this tag in chat also after this post, but there was not much discussion. Mar 30 at 9:50
• I have recently come to appreciate these matrices, and am very happy for the list here Nov 27 at 12:15
• @CalvinKhor If I may ask, what led to your recent appreciation for rank-$1$ matrices? Nov 27 at 12:33
• its for an application in theoretical fluid mechanics. it is a lemma of Nash (see Nash-Kuiper theorem) that for any compact set $S$ in the space of postive definite matrices, you can find a finite collection of rank-1 matrices $kk^T$ such that they reproduce any $A\in S$ in the sense that $$A = \sum_k c_k(A) kk^T$$ with positive coefficients. Some smart folks figured how to use this in conjunction with a large collection of exact solutions to Euler to show non-uniqueness of (sufficiently) weak solutions (and later Navier--Stokes, but not the sort as claimed in the Millenium problem) Nov 27 at 12:42
• @CalvinKhor Thank you for the reply. I believe I started to appreciate rank-$1$ matrices when I read Vinzant's What is... a Spectrahedron? [PDF]. Nov 27 at 13:08