Fighting Fire With Fire : Crowdsourcing Security Solutions on the Social Web. Christo Wilson Northeastern University [email protected] High Quality Sybils and Spam. FAKE. We tend to think of spam as “low quality” What about high quality spam and Sybils ?. Christo Wilson.
Fighting Fire With Fire:Crowdsourcing Security Solutions on the Social Web
We tend to think of spam as “low quality”
What about high quality spam and Sybils?
MaxGentlemanis the bestest male enhancement system avalable. http://cid-ce6ec5.space.live.com/
Also used by spammers
Real or fake?
Screenshot of Profile
(Links Cannot be Clicked)
Testers may skip around and revisit profiles
More Profiles for Experts
Data from Renren
Not so good :(
80% of experts have >90% accuracy!
Almost Zero False Positives
Experts Perform Okay
Turkers Miss Lots of Sybils
Treat each classification by each tester as a vote
Majority makes final decision
Most workers are >40% accurate
From 60% to 10% False Negatives
Filter Out Inaccurate Turkers
Maximize Usefulness of High Accuracy Turkers
All testers speed up over time
Experts perform well on most difficult Sybils
Really difficult profiles
Showing profiles to crowdworkers raises privacy issues
Solution: reveal profile information in context
Public Profile Information
Friend-Only Profile Information