Conversation started Feb 4, 2017 at 17:06.
M.A.R.
M.A.R.
Feb 4, 2017 17:06
Guyses, what are some sites with a good tag system?
I want to analyze stuff.
I'm looking for sites that have kept a clean tagging system. Without meta tags, bad tags, or jack-of-all-trades tags
user315433
Drupal Answers
user315433
Not kidding. Compare the counts of questions and tags in this table: data.stackexchange.com
user315433
The sites are ordered by the number of questions. Reading the tag count column shows which sites have unusually low or unusually high number of tags.
M.A.R.
M.A.R.
@zaq Which is a good thing? A high number?
user315433
Since you are asking for sites "without A, B, C" it looks like you want fewer tags.
M.A.R.
M.A.R.
Feb 4, 2017 17:11
I do, but it's because I wanna cull tags
ELU has only 929 tags and I'm damn sure that site's tags are broken
user315433
Okay, so it's not about quantity so much but about too broad/vague tags.
user315433
Like single-word-requests or such
M.A.R.
M.A.R.
Exactly
My theory is that a good tagging system has some, say 5-10, really broad tags but tags you can quickly choose from.
It also should have very few tags with only one or two applications
user315433
Biology has 683 tags vs 253 on Chemistry, with fewer questions overall.
user315433
But this is really subject specific. Programming sites are naturally divided by language, that's an obvious choice for primary tags.
M.A.R.
M.A.R.
Feb 4, 2017 17:16
Only 14k questions? Dang, they really fell behind
user315433
In Sciences... can one really split organic-chemistry into subfields in a useful way?
M.A.R.
M.A.R.
@zaq Yeah sure, but each of those are tags as well
user315433
By splitting I meant you wouldn't need the umbrella tag any more.
M.A.R.
M.A.R.
@zaq At first glance, the systems look alike. 5-10 really broad tags that come to mind first when you wanna choose tags. And a mass of moderately sized tags
@zaq It's too useful not to have
user315433
On Math site, there are popular tags real-analysis, complex-analysis, functional-analysis, harmonic-analysis, fourier-analysis, but the umbrella "analysis" tag is discouraged as not useful. (And periodically proposed for burnination.) Again, subject-specific...
user315433
Feb 4, 2017 17:20
Tags categories: fields of expertise > objects of study > properties of those objects
M.A.R.
M.A.R.
An organic chemistry question is either about reaction mechanisms, kinetics, thermodynamics, a specific reaction or IUPAC nomenclature.
None of these can indicate the organic-chemistry-ness of the question well.
Frankly, it sometimes even feels under-applied
Is there an easy way to fetch a list of tags, a list that I can modify, like in Excel?
user315433
Together with their counts?
M.A.R.
M.A.R.
That would be nice to have, but not necessary
Using SEDE sounded a bit messy IIRC
user315433
I don't see why it's messy: data.stackexchange.com/chemistry/query/623602/list-of-tags you get a CSV file, import into Excel / Google Sheets/ etc.
user315433
TIL that the TagName field magically linkifies in SEDE
M.A.R.
M.A.R.
Feb 4, 2017 17:27
Hmm, let's see what I can do
I'm not actually trying to fix Chem. I'm trying to fix ELL
The thing that stinks in Chem right now is the HW policy, and people seem to keep misunderstanding the new post I wrote about it. :/
user315433
Feb 4, 2017 17:38
I was not fond of that post myself, and Oprah meme didn't help.
M.A.R.
M.A.R.
@zaq I've not posted for such a long time that I have lost my touch
user315433
Why should I care what someone else is interested in? What matters is if I'm interested in the question.
M.A.R.
M.A.R.
Probably
But I should first understand what they misunderstood in my post
@zaq Hmm?
user315433
> Textbook questions should demonstrate interest in the thought process leading to the answer, either by showing effort or inquiring conceptually different questions.
user315433
I don't care what the OP is interested in.
M.A.R.
M.A.R.
Feb 4, 2017 17:41
Well, that's not what I meant.
I meant something like 4.
Sometimes, it's obvious with the way the question is asked, that the OP knows what they're talking about
Throwing a boilerplate response and telling them to read stuff they already know isn't gonna help then
user315433
Anyway, the problem is unsolvable so you can just as well give up.
M.A.R.
M.A.R.
Thank you for your inspiration
user315433
The "effort/interest/conceptial" is a proxy for "I don't want to see the same crappy questions every day that aren't even questions but desperate pleas".
user315433
To which the answer is : "read a book instead".
M.A.R.
M.A.R.
Not for all users.
Thing is, it doesn't make much sense to just require effort in every question and close it if it doesn't comply
That's not how everyday conversations work
And I don't want Chem to turn into SO this soon
user315433
Feb 4, 2017 17:44
Agreed.
M.A.R.
M.A.R.
So I'm giving a reason to the same closers not to close a question they believe will invoke some good discussion
The same closers.
So this won't help the normal HW vampires
user315433
So, (1) allow interesting questions regardless of effort; (2) require effort for non-interesting questions.
user315433
Not optimistic about the enforcement of such a policy.
M.A.R.
M.A.R.
Well, more people than before would go home happy
user315433
This is where the words "elitist" and "double standard" come in.
user315433
Feb 4, 2017 17:49
And apparently, "elitist" is one step from "Nazi".
M.A.R.
M.A.R.
@zaq Point
Lemme think
Dang, this was such a nice idea
Now it looks like poop
user315433
So, read a book instead.
M.A.R.
M.A.R.
Hmm, I was trying to make an exception a rule.
Now I get it.
 
Conversation ended Feb 4, 2017 at 17:52.