2 New Curator Tools for Steemit: Green-tagging & Red-tagging

in #steemit8 years ago (edited)

Accurate tagging is of the highest importance for SteemIt as a content platform, especially as the amount of user articles grows. Users need a reliable way to filter through all the posts available to see what they want, and avoid having to see what they don't care about.


Currently, tags are added solely by post authors and are assumed to be in good faith, accurate and complete. I'm proposing a simple way to improve the tag system that completely removes these assumptions, allowing the curation community to vouch, and even collectively add to the existing author tags of a Steem article. Steemit can accomplish this by implementing two new mechanisms, which I called Green-tagging and Red-tagging.


Each existing tag displayed at the bottom of an article (as originally added by the author) becomes clickable, similar to the upvote button. When a curator clicks on a tag, this vouch-vote for the tag is recorded and the ordinarily gray tag turns a slight shade of green:

The more users click the tag (thus confirming or vouching that it is accurate) the more green the tag becomes. Each vote is recorded and curation rewards are given out for early accurate Green-tagging, just as they are for article upvoting.

  • Each curator can only green-tag up to 3 tags per article. Curation rewards are split between the three, so it may be wiser to green-tag just one tag that very accurately describes the article, than going for three random ones.


Where green-tagging works on existing (author provided) tags, the Red-tagging feature allows community curators to propose their own actual tags, if they feel the author was not honest or forthcoming about the true content of their article, or that the author forgot to include a really good tag that the article deserves.

We could represent the Red-tag ability as a little red +sign at the end of the author-provided tags display.

Clicking this + sign enables Red-tagging - it adds a new tag to the author's existing list, faintly outlined in red. The curator adding the red tag decides what the new tag is going to say. These tags must then be clicked on by other curators within some time frame, perhaps 2 or 4 hours, otherwise they will expire and disappear automatically.

This way we create a situation where an angry curator might add a tag such as #shitpost to describe an article that they feel is not of the highest quality :

Naturally the original author would never under normal circumstances include such a tag in their own post, because they know many users may choose to filter out #shitpost or #travelboobs globally from their Steemit experience. If many curators agree and "vouch" the red-tag someone added earlier (by simply clicking it), it gains staying power in increments, effectively becoming permanent and there is nothing the original article author can do about it.

Curators who add accurate Red-tags or upvote accurate red-tags early earn curation rewards from this activity.

  • I suggest allowing only one red-tag per curator per article to exclude the possibility of malicious red-tag spam.


More accurate, curator-vouched tags mean more accurate filtering on all of SteemIt. A typical scenario would involve a user going to their "filter" tab and getting a list of ~20 of the currenly most popular tags on the site. User can then click thumbs up or thumbs down on any displayed tag or search for more obscure tags manually and thumbsUp/Down for those.

These decisions are recorded per-user. SteemIt then uses the number of Green-tag votes associated with the particular tag in an article such as #music to push everything with that tag up in rank in that particular user's article list. The bonus ranking amount that an article gets in the user's SteemIt experience from accumulating green-tags is directly proportional to the current number of those green-tags, i.e. how many people saw the tag as being accurate for the post and clicked the tag.

This would mean that the more curators have approved a certain tag as accurate, the more effect it has on improving the ranking of the article for users who have chosen to thumbs up or thumbs down that particular tag. Imagine users being able to thumbs up what they want to see and thumbs down #begging or #shitpost, which we might imagine as popular red-tags.

Conversely, a user who thumbs down #musicvideo would see all articles that are green-tagged or red-tagged with that particular tag pushed down in their personal article ranking screen, again proportional to the amount of curator clicks for that tag. So, in a very intuitive way, these curator clicks can both add rank and take it away, depending on a user's individual preference.

In this way, we can see that red-tags are not always bad and can indeed be used successfully by curators to augment the author's tags (maybe they forgot an important one), earn rewards, and ultimately improve the post's ranking for those users who are interested in that particular tag.


We went over a few new ideas related to Green-tagging and Red-tagging a Steemit article. The two new features, although simple, work in synergy to empower Steemit curators in the crucially important area of article tagging. Accuracy and filtering improvements site-wide would begin to be felt almost immediately by all users, improving the Steemit experience, boosting its capacities as the flagship content platform and rewarding people who look after accurate article tagging, including the original article authors.

#steemit #steem #tags #curation #suggestion

Let's discuss it! Opinions and use-case scenarios are very welcome, as it's still an infant idea...

  • Update: User @tinfoilfedora contributed a re-tagging idea to make what I call Red-tagging in the article even easier. The suggestion is to simply be able to reply to an article with a newtag as the body of the reply. This quickly adds the new red tag, without having to click + buttons. Thanks @tinfoilfedora!

Great idea! This would work perfect with re-tagging.
If curators could add the re-tags in their replies, then your idea would work great. It fits right in with the other 2 features needed, user blocklists and tag blocklists. More details on that proposal here: Allow the users to fix the problem through curation.

Thanks. I'm not too familiar with re-tagging, how does that work?

It should work by simply adding #newtag to a reply.
Combine that with your idea, and this place would be easily organized by user opinion in just a few replies. Once users can block tags they don't like, everything offensive would just disappear.
Hopefully they will do something quickly.

Awesome. Never thought about using the Reply mechanic for it, that makes it effortless. Don't have to hunt with the mouse for any + signs.

Sounds like a good idea, but would need to sit with it. I love that people are so involved in the evolution of the platform!

The concept is great , think you will get a lot of supporters , good luck!

Cheers, I just hope the developers eventually read it.

Interesting idea, and good writting. Your post might not get the expected reward for the time you spent on it but i give you my upvote.

Glad you like. I'm happy as long as it gets seen by the devs or someone whispers in their ear about it.

This is an excellent idea and hope the devs will take a look at it considering the issues we are seeing with tag abuse or posts that are wrongly tagged.

Great post and fantastic ideas! I've been thinking a lot about tags and how steemit handles them and these changes would make a huge impact on serving the best content to the right people.

One question for you, regarding the following bullet:

  • Each curator can only green-tag up to 3 tags per article. Curation rewards are split between the three, so it may be wiser to green-tag just one tag that very accurately describes the article, than going for three random ones.

Wouldn't you want to gather as much information from curators as possible, instead of incentivizing "strategic" voting?

For example, a post about Pokemon GO would be relevant content for the "pokemongo" "pokemon" and "gaming" tags, at minimum. But the most relevant tag is "pokemongo" so all the savvy users would upvote only that tag, leaving the content unavailable to users who haven't expressed explicit interest in "pokemongo" but might be interested based on their interest in "pokemon" and "gaming".

Sorry if I missed something that counteracts this incentivized behavior, looking forward to your reply. Thanks again, it was hard to find anything I didn't immediately agree with. :)

Thanks and great question.

I did not go into detail on this point, but the way I see it, the purpose of Green-tagging is to sort of highlight the most relevant few of an author's original tags. So all the original tags are still there, but curators decide which ones they think are most relevant. For some articles, that might be just one tag that almost everyone will green-tag. Remember that the green-tagging influences rank based on a custom per-user filter (if they chose to set it up) on SteemIt. It doesn't cancel out or overpower the original tags from the author, no matter how many green-tags some tag gets, the article should still be searchable using the original tags...

For some articles, like in your pokemon example, it might actually pay more to green-tag three different tags: since other curators are likely going to feel these are the most relevant too, and your curation reward is split between the 3 but ends up being greater overall, thanks to high numbers of green-tagging on all three tags you picked.

I think we have to limit the number of green-tags because if we don't, then bots will blindly upvote everything on everything, in an effort to grab curation rewards unfairly.

I think this idea is just absolutely genius. The only issue would be if a couple curators come by and vote unjustly...which could happen. For the most part this would help with tag relevancy.

That's very true, but Steem on the whole is quite defenseless against brigades of that type, even with the system we have right now. Ways to combat brigading might have to be copied from Reddit or totally reinvented for Steem, something beyond this humble tagging article.

Unfortunately, this opens the door to censorship by proxy. Like the Hugo Awards "puppies," it enables organized factions to downvote opinions they disagree with, burying them. One red tag per curator per article doesn't prevent this; it only doubles the attack potential: bad actors could downvote both the article and the tag.

So far, downvoting has been contained to effective and honest filtering. I doubt that will be true going forward as the community swells and system manipulators get savvy.

Worried about brigades too... but it's a pretty general worry, we know there's nothing really stopping the same brigades from using the existing downvote (or upvote) for their nefarious SteemIt-gaming purposes.

What encourages me is that down the line (with millions of subscribers) although the brigades might be able to censor dissenting minority opinions off of "trending" or effectively "the main page", even then, users who are specifically interested in certain tags should still see these particular subjects they want rising to the top of their private main page list, if green-tagging and red-tagging are in place, doing their job.

everybody can choose what have value for them. stigmatizing is not good direction.

Coin Marketplace

STEEM 0.26
TRX 0.11
JST 0.035
BTC 65352.49
ETH 3165.93
USDT 1.00
SBD 4.06