Zhao Yibo Zhu Manish Mohanlal Haitao Zheng and Ben Y Zhao Computer Science Department UC Santa Barbara Serf and Turf Crowdturfing for Fun and Profit Review posted on Yelp ID: 635400
Download Presentation The PPT/PDF document "Gang Wang , Christo Wilson, Xiaohan" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Gang Wang, Christo Wilson, Xiaohan Zhao, Yibo Zhu, Manish Mohanlal, Haitao Zheng and Ben Y. ZhaoComputer Science Department, UC Santa Barbara
Serf
and Turf:
Crowdturfing
for Fun and Profit Slide2
Review posted on YelpDetailed contentEven has a personal touchFacebook profile Complete informationLots of friendsEven marriedOnline Spam Today1
Stock Picture
FAKE
Been B.
West Lafayette
IN, USA
Great
lyonnese
food: the "
saucisson
pistaché
" is delicious
.
Awesome
athmosphere
:
everytime
someone has his/her birthday, they turn the lights off and play "Happy birthday to you" while a waiter brings the
birtday
boy/girl an "
omelette
norvegienne
".
Reviews
for
Brasserie Georges
FAKE
High quality fake reviews and fake accounts!Slide3
Variety of CAPTCHA testsRead fuzzy text, solve logic questionsRotate images to natural orientationIdentify friends (Social CAPTCHA)Detectors using behavioral modelsDetect bursts in per-IP application requestsDetect bursts of new accountsSynchronized traffic from groups of accountsDefending Automated SpamRotate below imagesWho is tagged in the photo?
But what if the enemy is a real human being?
2Slide4
Black Market CrowdsourcingOnline crowdsourcing (Amazon Mechanical Turk)Admins remove spammy jobsNEW: Black market crowdsourcing sitesMalicious content generated/spread by real-usersFake reviews, false ad., rumors, etc. 3
Crowdsourcing + Astroturfing = CrowdturfingSlide5
Biggest dairy company in China (Mengniu)Defame its competitorsHire Internet users to spread false storiesImpact Victim company (Shengyuan)Stock fell by 35.44%Revenue loss: $300 million National panic4“Dairy giant Mengniu in smear scandal”
Real-world Crowdturfing
Warning: Company Y’s baby formula contains dangerous hormones!
MSlide6
Questions Asked in Our Study…How does crowdturfing work?Measure 2 largest crowdturfing sitesAnalyze growth, economics, workers, etc.How effective is crowdturfing?Infiltrate the systemPerform benign end-to-end experimentWhat is next for Crowdturfing?Crowdturfing in US and elsewhere Defending against crowdturfers5Slide7
OutlineIntroductionCrowdturfing in ChinaEnd-to-end ExperimentsWhat’s Next6Slide8
Crowdturfing SitesFocus on the two largest sitesZhubajie (ZBJ)Sandaha (SDH)Crawling ZBJ and SDHDetails are completely openComplete campaign history since going onlineZBJ 5-year history SDH 2-year history7Slide9
Worker Y ZBJ/SDH
Crowdturfing
Workflow
Customers
Initiate campaignsMay be legitimate businessesAgents
Manage campaigns and workers
Verify completed tasks
Workers
Complete tasks for money
Control Sybils on other websites
Campaign
Tasks
Reports
8
Company XSlide10
9Report generated by workersCampaign Information
Get the
Job
Submit Report
Check DetailsCampaign IDInput MoneyRewards
100
tasks, each
￥0.8
77
submissions accepted
Still need
23
more
Promote our product using your blog
Category
Blog
Promtion
Status
Ongoing
(
177
reports submitted)
URL
Screenshot
WorkerID
Experience
Reputation
Report ID
Report Cheating
Accepted!Slide11
SiteActiveSinceTotalCampaignsWorkersReports$ forWorkers$ forSiteZBJNov. 200676K169K6.3M$2.4M$595K
Jan. 08
Jan. 09
Jan. 10
Jan. 11ZBJSDHCampaigns$
Campaigns
$
High Level
Statistics
10
1,000,000
100,000
10,000
1,000
10,000
1,000
receptif2.package@gl-events.comSlide12
Spam Per Worker 11ZBJSDH
Prolific workers
Large number of transient workers
Transient workers
Makes up majority of a diverse worker populationProlific workersMajor force of spam generation Slide13
Are Workers Real People?12Late Night/Early MorningWork Day/Evening
Lunch
Dinner
ZBJ
SDHSlide14
Campaign Target# of Campaigns$ per Campaign$ per SpamMonthly GrowthAccount Registration29,413$71$0.3516%Forums17,753$16$0.2719%Instant Message Groups12,969
$15$0.7017%
Microblogs (e.g. Twitter/
Weibo)4061$12$0.18
47%Blogs3067$12$0.2320%Top 5 Campaign Types on ZBJ
Most
campaigns
are spam generation
Highest growth category
is
microblogging
Weibo: increased by 300% (200 million users)
in a single year (2011)
$
100
audience of 100K
Weibo
users
Campaign Types
13Slide15
OutlineIntroductionCrowdturfing in ChinaEnd-to-end ExperimentsWhat’s Next14Slide16
How Effective Is Crowdturfing?What is missing? Understanding end-to-end impact of CrowdturfingInitiate campaigns as customer4 benign ad campaigns iPhone Store, Travel Agent, Raffle, Ocean Park Ask workers to promote products15
Clicks?Slide17
Weibo (microblog)End-to-end Experiment
Measurement
Server
Create Spam
16
Travel Agent
Redirection
Campaign1: promote a Travel Agent
New Job Here!
ZBJ (Crowdturfing Site)
Workers
Task
Info
Trip Info
Great deal!
Trip to
Maldives!
Check Details
Weibo UsersSlide18
Campaign ResultsCampaignAboutTargetInput$Task/Report
Clicks
Resp. Time
TripAdvertise for a trip organized by travel agent
Weibo$15100/108283hr
QQ
$15
100/118
187
4hr
Forums
$15
100/123
3
4hr
17
Settings:
One-
week
Campaigns
$
45 per
Campaign ($15 per
target)
Cost per click (CPC)
Weibo
($0.21), QQ ($0.09
), Forum ($0.9)
Price > Web display Ads ($0.01
)
80% of reports are
generated in the first
few hours
receptif2.package@gl-events.com
receptif2.package@gl-events.com
Averaged
2
sales/month before campaign
11 sales in 24 hours after campaign
Each trip sells for $1500Slide19
OutlineIntroductionCrowdturfing in ChinaEnd-to-end ExperimentWhat’s Next18Slide20
Crowdturfing in USGrowing problem in USMore black market sites popping upInternational workers who speak EnglishSites% CrowdturfingMinuteWorkers70%MyEasyTasks83%Microworkers89%ShortTasks95%19Slide21
Where Is Crowdturfing Going?Growing awareness and pressure on crowdturfing Government intervention in ChinaResearchers and media following our studyCrowdturfing sites will respond and adaptHide campaign details/historyMigrate to private communication channels20
Defending against
Crowdturfing will be very challenging!!Slide22
Ongoing Work: DefensesInfiltrate and disruptMasquerade as bad customers or workersOverwhelm the verifier with floods of bad reportsDetection using statistical modelsIdentify patterns of workers and campaignsTemporal behavior models21Slide23
ConclusionIdentified a new threat: CrowdturfingGrowing exponentially in both size and revenue in ChinaStart to grow in US and other countriesDetailed measurements of Crowdturfing systems End-to-end measurements from campaign to click-throughsGained knowledge of social spams from the insideOngoing research focused on defense22Slide24
Thank you!Questions?