Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Massive networks of fake accounts found on Twitter

bbc.com

Post reply on HN
Print view View original

141–150 of 267 posts

Re: Massive networks of fake accounts found on Twitter

#141

Post by KirinDave »

I'm just amazed at this entire series of threads. "Gosh, who knew violating social network ToS by proxy and creating massive false consensus networks was so easy?"

Everyone. Everyone should know.

We're watching this terrible trend rip apart the entire social proposition of the internet after spending 2 decades trying and finally achieving buy in. And here y'all are, hopefulls for a digital economy cheerfully defrauding the very networks that will probably bet the monetization strategy for many startups that pass through HN's doors.

The total lack of any personal responsibility here, or notion of consequence... It stuns me.

Re: Massive networks of fake accounts found on Twitter

#142

Post by stephengillie »

aresant wrote: post #55
jashkenas wrote: post #28

I would imagine that it's possible that some of the people reading these comments might already know a lot more detail about Twitter/Facebook bots than this story goes into. Although you fine folks might remember me from a few other projects, I'm currently a Graphics Editor at The New York Times. We're actively curious and interested in pursuing lines of inquiry about this sort of behavior, and if any HN'ers have any…

Could NYT quantify - through experimentation - the value of a bot army as the "story"? For instance a Tweet with 100 retweets is substantially more likely to generate additional retweets / be shown more by Twitter's algorithm / etc vs. a tweet with 10 retweets. And scale that up! Easy to build an experiment and quantify even on a small level for NYT.

They could have tweeted, paid for a bot army to retweet, then reported on that.

Re: Massive networks of fake accounts found on Twitter

#143

Post by geofft »

dgfgfdagasdfgfa wrote: post #109
geofft wrote: post #29

Earlier quoted context omitted.

"Fake" implies deception: it's fake because it counterfeits or forges something else that is real. An anonymous or pseudonymous Twitter account run by a human, in the way that a human is expected to use a Twitter account, is not a fake account: it's real, just pseudonymous. A bot account that's clearly a bot, like @big_ben_clock or @choochoobot, isn't a fake account either: it doesn't pretend to be anything other tha…

Deception is a great feature. Why is that bad? I don't want Twitter to know my name. Don't exclude me for valuing my own privacy! I'm certainly not "fake", and neither are my interactions.

I'm not sure how you got that interpretation of "deception" out of what I said. You're not fake, and I'm not excluding you. You're just also not deceiving. I certainly don't think your real name is Dgfgfdagasdfgfa.

To repeat what I posted: An anonymous or pseudonymous Twitter account run by a human, in the way that a human is expected to use a Twitter account, is not a fake account: it's real, just pseudonymous.

Re: Massive networks of fake accounts found on Twitter

#144

Post by throwaway112991 »

this is relevant is as well: https://play.google.com/store/apps/details?id=com.hitwe.andr... tldr: 'HitWe' it's a tinder-like dating app with 10M+ installs on google play (haven't checked itunes). i came across it by browsing dating apps for fun. most of the bad reviews (1-2 stars) are from people who claim it's 95% fake profiles etc. many claim that women there just ask for their email etc. some even claim to have paid some of the women which have then disappeared (but that's being too naive, IMO. never pay someone you have never met.. ). there are many many good reviews, but most are only with 5 stars and without any text (which, to me, is another red flag).

anyway - the most interesting part here is that they actually managed to fool google, up until this very moment! google recommended me to view their app.. spam does work !

ps - googling (ironic, i know) 'hitwe app scam' showed me this on the first result: http://www.datingbusters.com/hitwe-com-exposed-for-fake-prof...

i am interested in a response from one of google play's spam engineers/managers ..

edit: spelling

edit2: it took me 0.5 secs to start sensing that it's a fake-boosted app. a human reviewer at google could have just scanned the top 100x dating apps in a single day and map out the fake apps. what do you think?

Re: Massive networks of fake accounts found on Twitter

#145

Post by megablast »

leephillips wrote: post #135
_pecl wrote: post #131

Earlier quoted context omitted.

yes. I'm not sure about all of the details; I'm sure there are a variety of indicators that go into spam score and once you tip the spam threshold you're banned. So I would guess that students on a school network do things well enough to stay safe, but if you have 1,000 bot accounts on the same IP it's just a matter of time before they're all gone. So if you want to manage a network of fake accounts I try to use a pr…

OK. I guess you could spoof your IPs in the bot program if you don't care about seeing the reply, which for this application you probably don't.

A little bit of knowledge....

Re: Massive networks of fake accounts found on Twitter

#146

Post by CharlesW »

pavanky wrote: post #95
astrodust wrote: post #47

Earlier quoted context omitted.

I think that's an optimistically low estimate.

Why ?

It's not possible to be completely sure that 5% isn't a reality-based estimate, but it's typical for services like Twitter Audit and BotOrNot to identify half or more of a celebrity's followers as bots.

For example, as I type this BotOrNot estimates that 34% of @RealDonaldTrump followers, and 48% of @BarackObama followers, are real.

Just from being a nobody on Twitter who uses it almost every day, I'd also agree that 5% is laughably optimistic.

Re: Massive networks of fake accounts found on Twitter

#147

Post by nickfromseattle »

problems wrote: post #89
andai wrote: post #53

Earlier quoted context omitted.

Would you mind sharing why you did it? I've been curious about this for a while.

The answer probably isn't satisfying - pranks on friends, it's cheap and is usually a good laugh. 20k followers for around $10 found on a blackhat SEO site. 19.5k of which are still around after about a year. You could totally use it for business purposes too - as a really shitty form of advertising or just to make your popularity look higher. I doubt it'd be too effective, but hey, it's cheap.

I'm familiar with this space and I think the biggest benefit is providing 'social proof' to new websites / startups / brands.

I know from my own experience when doing research on a company - whether it's a prospective client, vendor or competitor, I'll click through to their social properties to gauge their traction.

I definitely look at their engagement ratios to determine whether their social following is organic or fake, but I think typical consumers miss this.

Re: Massive networks of fake accounts found on Twitter

#148

Post by groby_b »

pfarnsworth wrote: post #104
throwaway91111 wrote: post #63

Earlier quoted context omitted.

Yes, but a number of difficulties remain: 1) there's little way of figureing out who is actually contacting you. 2) If you file the number with the FCC you don't get a response. 3) it's not always obvious there's a commercial entity on the other side. I've gotten a number of calls that appear to try and keep me on the line while preventing from Turing test esque conversation from taking place. 4) some of the spam is…

For #3, I no longer answer the phone with "Hello" from an unknown number. I wait for the person to say "Hello?" or if the person starts talking I just hang up.

All unknown numbers go to voicemail. Spammers inevitably leave 2-3 seconds of silence, so I just ban based on message duration.

Usually, I report about 5-10 numbers a day.

Re: Massive networks of fake accounts found on Twitter

#149

Post by swiley »

ChuckMcM wrote: post #117

The question this sort of thing leaves me with is "Why does {Twitter/Facebook/Google} permit this? What are they getting out of it?" Its pretty clear what the bot guys get out of it, pay to promote services, pay for followers, etc. They can monetize "fame" through the robotic horde. But as this article and ones before it point out, these networks are generally quite easy to spot. So why not take them out? It probably…

People who make money by being popular probably like it.

Re: Massive networks of fake accounts found on Twitter

#150

Post by ikeboy »

throwaway91111 wrote: post #63
dhimes wrote: post #51

Earlier quoted context omitted.

In the US isn't there a law expressly prohibiting these calls to cell phones?

Yes, but a number of difficulties remain: 1) there's little way of figureing out who is actually contacting you. 2) If you file the number with the FCC you don't get a response. 3) it's not always obvious there's a commercial entity on the other side. I've gotten a number of calls that appear to try and keep me on the line while preventing from Turing test esque conversation from taking place. 4) some of the spam is…

We need a service that will collect complaints, sue on our behalf and send a check, taking their cut.

Have an app that allows you to report a phone call quickly, automatically recording and uploading the entire call (it should record all calls and delete recordings after a few hours if not reported). Then for numbers that have been reported by multiple people, have someone investigate, and if the source is located, file lawsuits.

It's $500 per violation, so if there's enough people lawsuits can work, and the company making the app can reduce investigation costs by distributing it among many different complainers (so even if a private investigator costs several thousand to get contact info, if they called 100 people and they end up getting a 50k judgement it works out).

Post reply on HN