Twitter analysis: identifying a propaganda bot network

https://www.bellingcat.com/news/2019/09/03/twitter-analysis-identifying-a-pro-indonesian-propaganda-bot-network/

by anigbrowl • 7 years ago

232 71 7 years ago

markdown• 7 years ago

Surprising to see this on here. The ongoing genocide of the West Papuan people by Indonesia has largely been ignored by the West.

BTW the propaganda team behind this bot network has also been very active on reddit.

nencrystation• 7 years ago

Every now and then bellingcat covers a "lesser" news story like this one, so they can pretend their main purpose isn't to push for regime change in Syria.

howard941• 7 years ago

bellingcat made its mark documenting the MH17 shootdown. https://www.bellingcat.com/tag/mh17/ has the exhaustive history

varjag4• 7 years ago

It's not just chemtruthers and Assadists that are mad on Bellingcat. They stepped on plenty of toes elsewhere.

peteretep2• 7 years ago

So?

adipginting3• 7 years ago

You need to check the definition of "genocide" before saying that it is on going. There are human right abuses and neglect by the central government. These used to be happened all over Indonesia and still are but not just in Papua. The government is trying to fix the situation. There will be no another independence to another province of Indonesia

guelo• 7 years ago

Are you a member of the Indonesian propaganda team?

anigbrowl• 7 years ago

Though I can't keep up with every political conflict I'm particularly interested in the intersection of genocide, propaganda, and social media. It's OK to gmail me with tips or questions.

eecc• 7 years ago

Uh, from the article: why is he using sudo to run a python script?

  sudo twint -u bellanow1 –media -o papuabots1 –csv

Not good

speps• 7 years ago

Theory: he's a reporter, someone showed him the command, one time it didn't work and the other person said to use "sudo" possibly because he was trying to write a file somewhere he needed access to. Now he always uses sudo because it always works...

_5mpi• 7 years ago

Because confusing state of version management in 2019: https://twitter.com/noneprivacy/status/1168955093934399493

If you used sudo because you installed the package "system wide", try with "pip3 install --user twint" which will install twint in your home directory, and so you'll not be asked to use sudo

vernie• 7 years ago

I do not like the color of this bikeshed.

cat1993• 7 years ago

because sudo improves security over 'su' and raw root logins, obviously

luckylion• 7 years ago

This exact sentence was published by many of the other bots in their own individual tweets using the hashtags #westpapuagenocide and #freewestpapua.

A clickbaity video "what the government is trying to hide" that contains pro-gov messages ("they've developed an app for personal finance management and are giving villages more opportunities", "they are helping the food industry") posted under those hashtags. Is that attempting to influence people that are looking up those hashtags (presumably expecting to find info on "the genocide"), or is that just to add noise to the hashtag?

It looks too high quality for just noise (so it's probably too expensive to produce a lot of that content), but on the other hand, I can't imagine someone saying "woah, there's a genocide going on in west papua? I better check twitter. Ah, here's a video ... mhh, they are deploying an app to help people with finances and there's something about the food industry? I'm now convinced there is no genocide". Like, is that some "if only people knew about our app, they wouldn't mind the crimes" thinking? Does that work?

Jsharm• 7 years ago

If it's possible to do this analysis why does twitter not routinely shutdown these networks?

scribu• 7 years ago

The analysis in this post is only on two rather small hashtags (a very small percentage of the whole of Twitter).

It also involves a lot of "eyeballing", particularly the part where the author looks at the graph and notices a weird pattern. That's very hard to do automatically.

molmalo• 7 years ago

This. And also, the cynic in me tells me that closing millions of accounts does not look good in your quarterly report on user base growth/active users.

brudgers• 7 years ago

Twitter could pay people to do the work.

Nextgrid2• 7 years ago

That's very hard to do automatically.

When you run a worldwide network that has significant potential for abuse and spreading misinformation I would consider it to be your duty to monitor it and guard against these kinds of things. If it costs too much then maybe you shouldn't be in business or should change your business model so that you can afford to clean up the damage your business is doing to society.

milofeynman• 7 years ago

That would affect their user numbers they give investors...

r721• 7 years ago

Sometimes they do: https://blog.twitter.com/en_us/topics/company/2019/info-ops-...

rc_kas• 7 years ago

not often enough.

wnmurphy• 7 years ago

I would guess it's traffic + engagement.

notadoc• 7 years ago

Do propaganda bots increase user engagement metrics?

spamizbad• 7 years ago

Twitter shuts them down when they reach the press. Journalists emailing them about botnets will get them shut down faster than actual security researchers filing detailed reports.

My theory: botnets that fly under the public's radar allow them to sandbag their engagement stats. Once said botnets become public knowledge, it becomes a liability to advertisers and users who now are going to be worried about spam, so the fear of net loss of engagement spurns them to take action.

peterangular3• 7 years ago

Same reason Facebook bends over backwards to make major figures in politics happy, or companies like Blizzard want to make certain governments happy, etc.

It's money. It's always money.

oarabbus_2• 7 years ago

Can you point to some examples of Facebook bending over backwards to make major figures in politics happy?

Something besides the Warren allegations; MSNBC, NBC, ABC, Twitter, YouTube, and many others also ran the ad.

Anon84• 7 years ago

More than possible, it's been relatively simple to do for practically a decade. Case in point, here's one of my papers looking at twitter bots from 2011: https://www.aaai.org/ocs/index.php/ICWSM/ICWSM11/paper/viewP...

For a recent (last week!) review checkout https://arxiv.org/abs/1910.01720 (by a former colleague)

davidw• 7 years ago

Facebook is similarly pretty bad: https://popular.info/p/the-tiny-alaskan-public-broadcaster-5...

jammygit• 7 years ago

On the other hand, my twitter account was banned almost immediately after creating it this year. They said it was for “breach of terms of service” - and they won’t even let me delete the empty account now and declared they won’t respond to my support requests.

I think it was my VPN. I just wanted to thank somebody for a helpful tweet

freeflight2• 7 years ago

Because the false positive rates are still rather high as there's no universal and foolproof definition of what behavior reliably tags people as bots.

ljw10012• 7 years ago

just like spam email, and yet hardly any get past Gmail's filters.

ljw10012• 7 years ago

Because all those bots look good on the recurring visitor stats and the provocation is good for engagement

ailideex2• 7 years ago

Not a bot, Twitter recently blocked my account that had almost no activity without providing any reason and now I cannot delete it which is in contravention of GDPR.

peteretep2• 7 years ago

I cannot delete it which is in contravention of GDPR

Go on. They’ve refused your removal request with appropriate ID sent to their office, or?

ailideex2• 7 years ago

They’ve refused your removal request with appropriate ID sent to their office, or?

I have sent my email address to their support system while being logged in and they did confirm it. So yes. And they have not even responded so clearly in contravention with GDPR.

Go on.

With?

megous• 7 years ago

Appropriate ID is username/password if real-life ID was not previously linked to an account directly.

Twitter analysis: identifying a propaganda bot network

Related Stories