Maybe this script by @neil can be adapted for spam detection: https://mastodon.scot/@neil@glasgow.social/109832617051005111

To distinguish spam from non spam.

Lots of small artists and business owners will post about their life, but occasionally share and promote their work. I don’t consider this spam.

Promoting your work twice every day, I'd consider spam. But just banning these accounts may be hard to police.

We should have a business flag that hides their posts from the main/public/hashtag timeline but is still available on a "Business Timeline." We'd welcome them, but filter them out of the home feed.

A federated community is only as good as the moderators who put in the work. I expect folks who host or join Pixelfed to share images aren't motivated to moderate content and delete spam accounts.
If you don't deal decisively with the spammers, they will consume you. Some sort of built-in spammer check is a good start. Account creation doesn't have to be immediate.
[Take this advice from someone who was on the Internet when it was just engineers, scientists, and academics].

Why not start with the business account flag and see how it goes? You could always try the nuclear option next if that doesn't work.

Seems to me that instances could have different policies w/r/t the business account flag, with some welcoming business, others hiding it, and still others choosing some moderate limitations. That choice seems in the spirit of the Fediverse.

I have more SEO spam businesses following me than real people. At the same time, they're not posting anything, so don't clog up the feed. However, I would like to be able to prune my followers, If I block someone, I'd like them to disappear from my followers. It's more a minor annoyance than anything serious right now.

I think a business account tag that’s rigorously enforced and blockable could be useful. We could add exceptions to the things we like and want to follow (like bookstores!).

I agree with others that manual account review is a good solution to combat unwanted accounts on an instance. Further, I don't see a problem with deleting an account if it never made any posts, after a reasonable length of time. I also think your idea in another post about letting users limit the number of posts they see from each account they follow in a given amount of time is good. Businesses will show up on fedi, leave it to instances and users to naturally defederate and block.

I don't see how spam should become business accounts. And I don't see why business accounts should have limited reach

Welcome to our world said every Mastodon admin. The first tactic was to review sign ups but that now hurts your placement on joinmastodon so I personally need to suspend about 20-30 signups a day because they're advertising bots. I'm not sure if Pixelfed has a captcha system but Mastodon desperately needs one.

Add a simple captcha, like I did 😛

I don’t think business accounts is a terrible idea. Nuking spammers will be a lot of extra work for admins, and adding a way for people to interact with companies can be a way to get even more users to do the switch.

I get that it’s drifting away from the core values of Pixelfed but maybe it’s a necessity for growth?

put them on their own instance

it seems there are people here completely against any commercial activity in pixelfed. Others are open to see some categories of sellers (artists, authors, photographers etc) that may use the platform to share their work. In that case, instead of having only one type of flag, maybe it could be nuanced. Users would choose to get only non-commercial content in their feeds, or accept commercial content from some categories of sellers ("arts", "music", etc) or accept everything.

Central service sounds risky and Not In The Spirit Of Things.

Is there a verification step required? And could you automate a "no verification within 2hrs, auto-delete. No account activity (post/follow) within the first 12hrs; auto delete"?

It depends on what do you define as spam‽ Is it all business accounts or just the usual scammy spam accounts who trick people into paying for fraudulent goods & services‽

- Give the user the chance to opt-out of this sort of thing? This way you don't discourage Big Money from existing on your platform, but the user doesn't have to be annoyed by it.

If someone creates an account and leave. It would nice if we send him a reminder to verify his account within a certain amount of time or we'll suspend his account.

Another option could be asking "why you're is joining" in the registration form. Then we can use do some automation to check if the answer is valid or not. Typically spammers are not going to explain or write a valid answer.

Some of my fav accounts on Instagram are organizations (various art collectives and activist groups) and businesses (like commercial galleries and interior decoration firms). Why block or devalue them?

the first option seems too far fetched, it centralizes control on account creation (if I understood your post correctly).

I wouldn't forward registrations to a central authority - i wouldn't want to and it'd also violate gdpr.

the advertising account-flag sounds interesting, although i'd most likely keep advertisements banned completly on .de. Not looking to fund other ppl ads with my / donation money…

Sometimes I'd like a shadowban option to keep advertisers to register new accounts when banned. Not sure if it'd be effective though

As others have said, why is a company intrinsically spam?

If we want the Fediverse to be healthy, it's going to necessarily have some business/corporate participation. That's *good*. We want the next "Wendys" to be part of it.

Clearly marking a business account is good, like Mastodon clearly marking Bots. Categorize correctly and let people follow what they want.

You shouldn't penalize an account "just" because it's a business account. That's unhealthy for the ecosystem.

Adding a "business account" flag sounds interesting. Those, who absolutely want to act commercially on Pixelfed would thereby

1. get the possibility to be generally present
2. can not clutter up the Fediverse

@fediversereport like the idea. Might need more time on the label, “business account” might make professional/freelance photogs using the platform in good faith confused about which set of rules their account plays by

nuke em

If a post gets flagged by x users as #advertising it should be officially labeled as advertising. If an account accrues y officially labeled posts in a given period, the account should be labeled a #spam account.

That's my current opinion on how to protect servers from the flood of corporate spam that is imminent while simultaneously protecting small #makers selling #crafts, etc.

Then again, protecting #search results from spam might require a heavier hand. #IMO

Nuclear.

Then nuclear again until they glow in the dark and you can use the little blue remnants to light parking lots.

(Any legitimization of advertising is The End. It may come quick or slow, but it's The End, advertising being inherently inimical to community.)

Agree it's better to allow a properly identified voice than to try to ban it But I'd broaden the designation. Rather than "business" flag, how about "commercial interest"? to include individual users who are trying to monetize their presence.