Spam on wiki

How should spam pages on the Reprap wiki be handled? I noticed a new spam page at [reprap.org] created by the new user MKWBeulah02 [reprap.org] .

I can't block the user or delete the page, but I can remove the spam content.

@b4ndit are you still about, can you ban user and remove the page?

It's a pity, but it seems that these recent user accounts look like spammers and all follow the same pattern: a stock phrase like "My name is [RANDOM_NAME_HERE] and I am [RANDOM_OCCUPATION_OR_HOBBY_HERE]":

[reprap.org]
[reprap.org]
[reprap.org]

And that last account also created another spam page: [reprap.org]

It seems the RepRap wiki is the target of some specific spammer who keeps creating accounts and wants to sell patio chairs...

Edited 1 time(s). Last edit at 07/20/2021 08:44AM by qrp-gaijin.

Since no one came forward to assist. I contacted Dr Adrian Bowyer. He has alerted the wiki admins.

Hopefully this will be taken care of shortly.

Unfortunately, a quick glance at the recent changes on the RepRap wiki indicates the problem is continuing -- two more new spam accounts, two more spam pages about patio furniture.

Perhaps it would be wise, at least for the time being, to implement harsher restrictions, such as preventing new users from creating new pages?

Hey guys, sorry I didn't see any of this. I haven't gotten any emails from the forum for a while, so didn't get notified (I'm rarely around the forum anymore to see the notifications otherwise).

The SPAM on the wiki is primarily handled by AbuseFilter (https://reprap.org/wiki/Special:AbuseFilter). I created all the rules you see on that page and that's what eliminates the vast majority of the SPAM before it's posted. It's been a while since I last touched it, so it's likely in need of another update. A lot of the SPAM pages I've personally seen recently do not seem to include links, so do not hit any of the filters which aggressively filter links from new accounts.

Check out rule #3 (https://reprap.org/wiki/Special:AbuseFilter/3) and if you see common phrases that the SPAM bots use (that are otherwise uncommonly used by humans), we can add them to the filter to block the SPAM before it's posted.

I subscribed to this forum topic, so hopefully I'll get some emails if you guys reply.

... so if a spammer creates a page without links and add them manually later, it will undergo the spam-filters?

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Not quite. I urge you to check out the logic for each rule if you're interested, they're surprisingly human readable.

Filter #8 specifically handles that condition where brand new users try to edit links into pages. There was a fleet of SPAM bots that would edit existing pages and just replace links or add links to unwanted external sites. This rule filters that behavior out.

For the most part, 99.9% of new accounts trying to post link-based SPAM will be stopped automatically. The vast majority of issues the wiki currently sees are bots that post pages with no links at all. For that behavior, I rely more on keyword blacklists, but that needs to be updated.

... I'm mainly handling the forums, not the wiki -- here the filters are different, but seems too much better, than some years before ... but not perfect, neither

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Just for fun, I have switched the Captcha over from ReCaptcha (entirely useless anymore) to a Captcha system that allows custom questions and answers. So to register an account, a user must answer a very simple question that they should know prior to joining the site or can simply research given 2 minutes of effort (i.e. Who is the creator of the RepRap Project? Hint: A*** B***)

I've used this Captcha system in the past and it works really well against bots that are not human assisted. Eventually they seem to break through, though. We'll see what happens this time.

... I've heard rumors, that first spammers preparing to use GPT-3 for solving complex text based cpatcha's ?!?

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

That would not surprise me at all. Machine learning and AI have progressed by leaps and bounds and are very user-friendly using Python libraries.

We're arriving at a point where Captchas and other "human tests" no longer work. Well, that's not entirely true. It will likely always be possible to devise a test that only a human could figure out, but the design of the test has to be such that it's accessible to the dumbest people. It's exceedingly difficult to come up with a test that any human can pass that a machine cannot.

For our wiki, I'm more concerned about people that do not have a strong grasp of the English language being able to read, interpret, and pass the custom captchas. That's why the questions I created try to be exceedingly simple to understand and answer.

... hmmm ... as the forum is multinational and polyglot, pure English tests really could be broblematic -- but then even iconic or bitmap-puzzles are sometimes "culture-specific" ... e.g. traffic signs or such ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Captcha-type tests are not very useful as AI's can solve them better than humans: [www.dailydot.com]

Vik :v)

... hmmm -- then, maybe, we alredy have some AI "users" here in the forum? ... could be, they'll behave better (more sound) than some "human" users too?

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

More creative spam techniques have appeared on the wiki which are difficult (for non-admins) to remove:

1. Uploading an image. There seems to be no way for normal users to delete the content. Example: [reprap.org]

2. Uploading the spammer's phone number into the page title, which then cannot be edited. Example: [reprap.org]

It's sad enough that there is so little activity on the wiki and the forums... but I hope we can at least stave off the spammers.

... hmmm ... can't edit the content, but no problem to delete the complete page ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]