Hashtags Instagram Won’t Let You Search | Hacker News Reader

Hashtags Instagram Won’t Let You Search

thedatapack.com

67 pointsmaryvillavon54 comments

This reminds me of Shutterstock's open-source List of Dirty Naughty Obscene and Otherwise Bad Words:

https://github.com/shutterstock/List-of-Dirty-Naughty-Obscen...

This wins repo of the day!

Having been on several projects that required such lists, I'm glad that such a repo exists, but be aware that it is just the tip of the iceberg. Stakeholders emerge from the woodwork: such and such on the delivery team had a terrible experience before so we should include these words, HR and marketing have company-specific lists to merge in, some of the producers have unsettlingly precise and revelatory requirements ...

This is the most beautiful commit message I've read in a long time.

https://github.com/shutterstock/List-of-Dirty-Naughty-Obscen...

Now what exactly is the thing it refers to?

An exceedingly overweight and possibly unclean individual, according to http://nb.urbandictionary.com/define.php?term=shit%20blimp

Interesting. "anal" is a bad word in English, but not in German. On the other hand, "naked" is a bad word in German, but not in English.

Maybe I should send a pull request...

It may that they did not add it for a reason also. There are a lot of edge cases because many potentially dirty concepts are made up of words that are not bad alone. For example a text can have both "girls" and "nude" in it without being vulgar, but if it has the phrase "nude girls" the chance for it being pornografic is much higher.

( Searchdaimon have done some research on this and have a list if anyone is intrested: https://github.com/searchdaimon/adult-words )

There is also the data analysis perspective: http://qr.ae/8W4Pz1 :)

黒人 (black person) is on the Japanese list. I wasn't aware that that was a naughty word.

I removed it: https://github.com/shutterstock/List-of-Dirty-Naughty-Obscen...

TIL the word "anilingus." Thanks!

Some of the finnish ones:

jätkä - meaning literally "dude" hatullinen - hatful lahtari - an outdated word; I doubt many young people know the meaning/context of it. It was an insulting way to call the people on the white side during our civil war in 1918 pehko - thick hair

And they took down that repo.

> #bi > #gays > #lesbian

Yet another example of mentions of queer identities being filed as pornography.

Like:

#italiano #kansas #kickoff

Which presumably show racism towards Italians, a hatred of the midwest, and a dislike of football?

Or maybe Instagram has an automated system which correlated hashtags with reported porn. Tags with greater than X% reported porn get banned.

If that's true (and we have no reason to think it is or isn't true), then the banned tags come out simply because the #bi, etc. people happen to use those tags more for porn than (presumably) #straight people.

I'm wary of waving the racism / sexism / whatism flag. We're not all racist. We're not all sexist.

> Which presumably show racism towards Italians, a hatred of the midwest, and a dislike of football?

If that was a consistent pattern, maybe, but I am yet to hear of it.

> Or maybe Instagram has an automated system which correlated hashtags with reported porn.

> I'm wary of waving the racism / sexism / whatism flag. We're not all racist. We're not all sexist.

Maybe it's automated, sure. Maybe it's not directly intended. Doesn't change the outcome.

Also, no, everyone is racist and sexist to varying extents, it's not a binary. Humans are imperfect.

>Also, no, everyone is racist and sexist to varying extents, it's not a binary. Humans are imperfect.

Not everyone and everything is racist and sexist, stop peddling this histrionic narrative. If it's automated, are you claiming that algorithms are racist or sexist? The outcome isn't shaming a sexual orientation any more than x + 3 = 9 is "six shaming."

Presumably the claim would be that it contributes to a phenomenon which is considered harmful, and that there is an obligation to have it not do that.

Now, I don't know that I agree with that claim (certainly not without reservations at least), but I think that the idea of an algorithm contributing to a harmful phenomenon, and that being bad, is not an obviously wrong idea.

I don't think your argument against theirs is a very good one. Something better than your x + 3 = 9 quip would be needed to refute it.

I'm saying this as someone who also disagrees with the worldview you are arguing against.

/epistemic virtue signaling

The (management or computer) algorithm is likely designed to cut down on pornography and/or "offensive content"; the correlation of the terms bi, lesbian, or gay as metadata to explicit images is likely to be large on a visual platform. If anything is offensive, it's the quickness to jump to "I'm offended by logical explanations and I refuse to come up with practical solutions." Which is to say, I believe that the claim was infantile and deliberatively obtuse. Obviously it wasn't meant as a slight against sexual identity and if you want things to be better, get to work and stop complaining (which is unlikely to happen because complaining is easy and fixing things is hard).

Sometimes a cute riposte or simple response is all that's needed to defuse absurdity. Debating with someone plainly calling everyone racist and sexist isn't likely to bear fruit.

I wish we could get to a state in which the professionally offended would stop wasting others' time and encouraging this doublespeak-nonsense game -- there's only so many minutes in the day and we both wasted daylight on this.

TazeTSchnitzel's right. For all we know, users reported posts from proud #bi, #gay, and #lesbian people because they wanted those people off Instagram, not because there was anything wrong with their content.

The result is the same: three very meaningful tags that can't be used.

Probably because people were using them to post/promote their pornography using terms they know people would search. Please find something more interesting to feign offense over.

I don't care for the cause, I care for the result. And I'm not so much claiming offence as I am pointing out that it's a troubling pattern.

It isn't troubling. Those tags are primarily used to promote pornography. It makes sense.

Moral outrage isn't justified in this case. Sorry.

Open any porn site and I'm sure you will find these tags as top categories. It's not Twitter deciding it's pornographic or not, but users that use them as such, so admins simply have to react (unless you want Twitter to be overflown with gay/lesbian/bi porn).

Sure, I understand why this kind of thing tends to happen, but the result is no less insidious for it.

>so admins simply have to react

They should react in a fashion that does not involve blocking hashtags with multiple meanings.

> #snapchat

Huh. I wonder if there is some competition going on here

I would assume a lot of people posting "snapchat leaked" type pics (you know which ones I mean) with the hashtag #snapchat. When all those pics get reported the hastags on them get temporarily banned.

That was what by far I found most interesting. Those could be easily labelled as genuinely anti-competitive practices - (i.e., it'd be if Google blocked "Bing" lookups, or vice versa (presuming it was manually added by their staff, rather than auto-triggered by a lot of adult content being flagged at the same time, and a heuristic firing off a ban-signal)). MS got into that whole DOJ trouble for bundling their own browser for way less than that.

#kik is also in that list

I understand censoring pointless things like "kansas" but "asiandick?" Seriously? They're going to penalize Human A&P students that need help just because the source material is Asian!? Or just the Art majors that could get by with a different theme?

Well, that was a pretty silly list. Most of the hashtags here seem to be pretty normal or harmless stuff, at least to anyone above the age of about six.

Also, not a fan of this being done because 'some' content under those tags was reported as inappropriate.

Some of these are just weird or very specific. I’d love to hear the reasons for them being banned.

> #kissing

> #newyears

> #citycentre

> #hornyyyyyyasf

> #saltwater

There's probably little human review here. Most likely, hashtags often used on photos reported by users are automatically flagged.

Also why ban only #russianmilf?

Because everyone knows #russianmilf are especially salacious! /s

First Google result for #hornyyyyyyasf is: https://www.instagram.com/explore/tags/hornyyyyyyasf/

Instagram, you're doing it wrong!

Dang.

A wild guess, but #saltwater might be in there because it contains twat, a nice example of the Scunthorpe problem: https://en.wikipedia.org/wiki/Scunthorpe_problem

#easter and #kansas also seem kind of weird to me.

(And pity the chefs that want to show off their #eggplant parmesean, I guess...)

Is there a way to turn the filters off? Otherwise how is this not censorship?

It is censorship, and Instagram is just fine doing it.

It's censorship, but so what? It's a private company. They're allowed to display or hide whatever they please.

#workflow ?!?

#whitegirl?

come on!

Could be because 'white girl' is also slang for cocaine.

This, also it is a term often used by escorts promoting their services on instagram.

> #italiano

Really?

Italiano looks PG at worst.

https://www.instagram.com/explore/tags/italiano/

Kansas is weird.

https://www.instagram.com/explore/tags/kansas/

“Recent posts from #kansas are currently hidden because the community has reported some content that may not meet Instagram’s community guidelines.”

You aren't seeing the presumably bad stuff when you go to /explore because it is being hidden.

Ah thanks. Now I'm more curious.

I had figured that was a reference to the bizarre "Dilbert 2" video on youtube that instagram deemed abusable.

Watch the video and you might understand why.

no emoji?

Right, I recall a story being run about the blocking of the Eggplant emoji as a search term on Instagram.

Reply on news.ycombinator.com