Discussion about "lazy" tagging and usage of tagme (Formerly: Reducing tagme postings)

I don't see any harm in creating detail_request, it's also useful for asking help with chaotic pictures full of things going on. But maybe a more general name, like tag_request since we're asking for tags to be added?

As for aliasing tagme to undertagged, I don't mind either, but who exactly is going to tag images with "undertagged"? Uploaders themselves? Somehow I doubt that, with the negative connotations we're going to give with the alias. They'll just leave the posts actually undertagged without adding undertagged...

Type-kun said:

I don't see any harm in creating detail_request, it's also useful for asking help with chaotic pictures full of things going on. But maybe a more general name, like tag_request since we're asking for tags to be added?

As for aliasing tagme to undertagged, I don't mind either, but who exactly is going to tag images with "undertagged"? Uploaders themselves? Somehow I doubt that, with the negative connotations we're going to give with the alias. They'll just leave the posts actually undertagged without adding undertagged...

I think it's more to encourage people not to use tagme willy-nilly. Right now, they're using it in the "I'm too lazy to tag properly" way, instead of the "I'm missing something, but I'm not sure what" way the wiki defines it.

By making it an alias to undertagged, we do one of two things:

1) for those who didn't know what it meant or those who did know better but we're lazy, it discourages them from using tagme and might get them to tag a bit more.

2) for those who were using it to get help and don't know enough to search or use the forum, it'll be more descriptive about the post and maybe get some people on tagging it.

I'd keep tag_request separate from tagme, to avoid people using it the way they use tagme now. And if there are people who see the undertagged tag and then walk away without tagging it properly, they can be censured.

It's not like tagme is really helping with removing undertagged posts now....

A detail_request seems fine, though the resulting tag will likely require continual cleanup of those that have had requests filled but the tag not removed.

A concern with renaming tagme to undertagged is users simply stop using it altogether and minimally tagging the post, enough to not break the rules, but certainly lower than it could be and then no convenient tag on it to immediately pull it up.

Anyways, I don't have any strong feeling for or against renaming tagme to undertagged. Can't know the actual impact until the change actually occurs, and the alias can always be reversed or removed if the outcome proves poorly.

If they'll undertag a post now (using tagme as the excuse), then we won't see a lot of changes from them after the alias goes through. And we already can search by tagcount - we can either have a report to see how many new posts are showing up daily with a tagcount below a certain threshold (5-7 is what I'd use) to assess the impact.

Right now, doing a tagcount search for 5 tags gets me 111 pages of results, and none of the results in the first page have tagme in them. Given that, I don't think there's going to be that much change if we alias it.

Right now, gentags:<5 -no_humans status:active has 9877 posts.
gentags:<7 -no_humans status:active has 69187 posts.

fossilnix said:

Right now, gentags:<5 -no_humans status:active has 9877 posts.
gentags:<7 -no_humans status:active has 69187 posts.

The first 20 results in both searches are under tagged and don't have tagme included.

I don't see tagme making that much of a difference in the number of undertagged posts, not when several searches have gotten similar results.

Well, I have no grudge against the detail request tag, but it's just pointless in my eyes.
When it gets populated by one person, then this person could simply add the the missing detail.
When someone uses it, they might have read this topic and other topics as well (How do I tag this). Pretty much: We already have this tag and that's this thread. So if it's used, I'd simply point the user to this thread or to the comment section below the post. Because this way, we know what detail is missing. Detail request doesn't specify it. Is it a specific hat, is it the lingerie, is it something inside a messy room etc.

About the tags:
I'd go even higher and search for gentags:<11 -no_humans status:active. So we'd get here every post with ten tags or less and every post with a human in it, has more than ten tags. Even monochrome ones where only the face is shown, if they are tagged properly. See post #102891. Without male_focus there'd still be 12 gentags, which is quite accurate.

I do like the idea of a detail_request tag, also how about something like detail_check tag too? For those cases when somebody is not sure about the use of some tags in a picture.

Well, the idea is perfect. It's just not possible to implement it in a sane way^^'.

Btw. about fossilnix' post:
There are 70k posts with less tags than 7 tags. Tagme had 53k when I started reducing it.
So the next goal would be reducing this amount to a good amount. Problem: This will take an extreme amount of time. Nearly alone, tagme took me around 4 months to get it down to zero (lol, no life :c).
Well, a proposition would be to reduce these undertagged posts as well, but it will take a huge amount of time. What about that in the first place?

Provence said:

Btw. about fossilnix' post:
There are 70k posts with less tags than 7 tags. Tagme had 53k when I started reducing it.
So the next goal would be reducing this amount to a good amount. Problem: This will take an extreme amount of time. Nearly alone, tagme took me around 4 months to get it down to zero (lol, no life :c).
Well, a proposition would be to reduce these undertagged posts as well, but it will take a huge amount of time. What about that in the first place?

That's a huge project that can be worked on. Would a report be possible to create that identifies the number of undertagged posts created a day, maybe sorted by user? If we target poor taggers for education and correction (to start), and then not approve undertagged posts by then to reduce their upload ratios, would that reduce the scale of the issue?

We can't get everyone, but if there's chronic undertagging uploaders, we should cull them some via corrections and education - and, if necessary, removal.

Jarlath said:

That's a huge project that can be worked on. Would a report be possible to create that identifies the number of undertagged posts created a day, maybe sorted by user? If we target poor taggers for education and correction (to start), and then not approve undertagged posts by then to reduce their upload ratios, would that reduce the scale of the issue?

We can't get everyone, but if there's chronic undertagging uploaders, we should cull them some via corrections and education - and, if necessary, removal.

Well, the best way to "create a report" is working on it and tag the images. This way, at least with tagme it was like this, one encounters most of the time the same users who are doing this. So you have here 2 in 1 process: Reducing the undertagged images and seeing who is doing this very often. And then create a record, or contact a mod^^.

Just throwing this idea out there, but what if the report listed the top users that tag the least amount, and the top users that tag the most amount. The first would allow education and corrections as has already been said. The latter would help identify users for potential positive records and also be examples to others on what good tagging looks like, given that they are tagging appropriately.

BrokenEagle98 said:

Just throwing this idea out there, but what if the report listed the top users that tag the least amount, and the top users that tag the most amount. The first would allow education and corrections as has already been said. The latter would help identify users for potential positive records and also be examples to others on what good tagging looks like, given that they are tagging appropriately.

That was my idea but expressed in a clearer fashion.

I just left a comment on post #2386567 because several of his posts have tagme despite having enough tags to not be severely undertagged (see also post #2386542 and post #2586535 for more examples).

You're right.
I've added some more text to the post, and leading to the Thread since that's the better place for this kind of tagging^^.

About that list...it's not a bad idea. But since we have so many uploaders, from normal user to Contributors, it might be hard to maintain this list, if it comes true. But if it's possible to keep an eye on it, it might help tracking down bad taggers. I'm just not sure if that's technically possible, so maybe asking @Toks and @Type-kun might help and also @NWF_Renim and @Hillside_Moose if there is need for it at all^^.

Meh, this is again at 85 posts. That's even more than the last weeks where the counter was only between 70-80. I really have no problems reducing 85 posts. This doesn't take one hour, but I still want to have opnions about users who are using this tag even when their posts are good tagged:
Should I write them a DM leading to the "How do I tag this" topic (I did that now sometimes, but there are still many users left^^). So I want to know if anyone agrees with this.

Provence said:

Meh, this is again at 85 posts. That's even more than the last weeks where the counter was only between 70-80. I really have no problems reducing 85 posts. This doesn't take one hour, but I still want to have opnions about users who are using this tag even when their posts are good tagged:
Should I write them a DM leading to the "How do I tag this" topic (I did that now sometimes, but there are still many users left^^). So I want to know if anyone agrees with this.

It's either that, or we make sure those posts get a comment about tagging and maybe a lower priority in the approval queue.

I hope there aren't too many other tagme deletions like this one out there...
http://danbooru.donmai.us/post_versions?search%5Bpost_id%5D=1557983

edit: just looked at that user's edit history... Oh dear...

Bringing this dead thread back to life...

Does anyone remember why the Uploader Bookmarklet lets people upload with no tags or maybe one tag again and slaps tagme on? I was under the impression that the Uploader Bookmarklet or the upload page in general refused to upload anything if you didn't have any tags.

Removing that "feature" might help matters, IMO. Granted, we're down to only 12 tagme posts now, but given the wiki for tagme specifically states that it's not an excuse to be lazy, having the Uploader automatically use it below a certain threshold of tags (2 in the case of post #2574317 when I first looked at it - the rest of the tags were only added about two minutes later) seems to defeat the point of stating that in the wiki.

Are you suggesting that there should be a minimum number of tags that have to be added to an image for it to be uploaded? Because if so the threshold would have to be really low to allow for the likes of post #1924241.

Does the Bookmarklet automatically add tagme? I'm someone who always tags their images before submitting so I wouldn't know personally but I've seen plenty of images get uploaded with one or two tags that haven't had tagme stuck on them. Are you sure it isn't just people manually adding the tag?

One tag.
But it doesn't really matter if we don't clean up the howto:tag wiki. One could say that it is a bit bloated.
Maybe I bring this up in the not so far future.