understanding Sybil attacks

[synctext]

This week feedback:

• 76 theory papers
• No real-world examples
• Buy them here, see study
• 8000 fake twitter example
• Perhaps buy accounts using Bitcoins and study them; ethics?

economics of the Sybil attack, using the back market insight it is possible to put a price tag and volume of market offerings for stolen websites, fake social network profiles, and fraudulent microblogging accounts.

[synctext]

TOC:

• "theoretical results" or theoretical foundations; with picture of trusted and non-trusted region, gullible nodes.
• "Sybils in social networking and microblogging" ? + picture of Twitter fake accounts + give sense of scale to readers (8000 fake ones)
• First a chapter with: understanding real-world Sybil attacks: with datasets, and pictures.
• "3. theoretical foundations", Clearly explain attack edge concept.
• "Sybil Attack" or 4. Sybil attacks in various applications
• "5. procuring Sybil attack accounts on darknets"
• buy accounts from 50 vendors and measure + dataset

Plans: (compiler constructions as course-load)

• hand-in for grading: 25 Nov.
• IEEE/ACM 2-column style
• rate control: n new identities/day per IPv4, few identities per hour for /16.
• added value of (a)symmentric reputation system
• mention dirty tricks attack: destroy reputation of a person ?
• MANETs and large-scale sensor networks only exist in scientific papers for past decades. So only minor mention.
• brainstorm: anonymity credits + privacy incentive + market thesis direction

[kc1212]

@synctext A few questions:

• Is it OK to use the ACM SIG Proceedings Template (see main.pdf)?
• What do you think of the following structure? I feel it's more natural to have the theoretical foundations in a separate section instead of keeping in the Real-world Sybil attacks part.
  • Introduction
  • Real-world Sybil attacks (the motivational part with pictures, give the reader a "sense of scale")
  • Theoretical foundations (attack edge model, etc.)
  • The Sybil attack in various applications (enumerate the various applications, P2P, reputation network, etc.)
  • Defense mechanisms
  • Related work
  • Conclusion
• How should I buy the accounts? SEOClerks and BoostLikes appear to be the popular ones. But the prior requires phone number verification and the latter only supports credit cards via PayPal. I'm not sure if I want to use my personal information for those.

[kc1212]

Recent example of Sybil attack - Twitter bots http://www.theatlantic.com/technology/archive/2016/11/election-bots/506072/

from the article: "more than a third of pro-Trump tweets and nearly a fifth of pro-Clinton tweets between the first and second debates came from automated accounts, which produced more than 1 million tweets in total"

more articles: http://politicalbots.org

[synctext]

To get the raw data.. Please try to email the 8000 Twitter people and the people within the YCombinator discussion thread that did follow-up analysis.

[synctext]

Great case-study: http://sadbottrue.com/article/68/

[synctext]

Also mention oldies

4.2.3 and 4.3.3 are more generic and not restricted to reputations systems. Mike Hearn provides facinating insight into real-world context of spam fighting, reputation mechanisms, and Sybils. "Worse, competitors could interfere with each others mail streams by submitting false reports. We see this sort of thing with AdWords." (2014)