Who is FAKE?
Discover Astroturfing or Attempts of Fake
Influence!
Lutz Finger Soumitra DuttaMiningData.biz
An Army of Bots
$ 2.76 million
The Social Net
The Social Net
The Analytics
The Problem
ANY measurement
(if useful)
Will be ATTACKED.
Mass Movement
•  Reach
•  Intention
•  Ease of Action
Mass Movement
•  Reach
•  Intention
•  Ease of Action
Mass Movement
Astroturf
•  Reach
•  Intention
•  Ease of Action
Bots 1.0 – Spammer going Social
@you malware.com
@you-as-well malware.com
D fresh-contact malware.com
Bots 1.0 – Spammer going Social
10.000 messages
in 4 month
@PeaceKaren_25	
  
Jacob	
  Ratkiewicz	
  et.	
  al	
  	
  -­‐	...
10.000 messages
in 4 month
Bots 1.0 – Spammer going Social
@PeaceKaren_25	
  
Jacob	
  Ratkiewicz	
  et.	
  al	
  	
  -­‐	...
Look for Un-Normal
Training:
•  @spam
•  Manual classification
•  Honey Pots
Look for Un-Normal
Differences:
•  Time: Regular or Bursty
•  Heavy Hashtag Usage
•  Blacklisted URL
•  Spam Words
•  Few ...
Thus Social Networks acted….
Thus Social Networks acted….
2013
7% SPAM
20122011
20% detection
Detected but still Dangerous
SMEAR Campaigns
Denial of
INFORMATION
Bot 2.0 – Social Bots
Analytics ToolsConversation Bots
More Human
Social Friend @JamesMTitus
More Human
Social Friend @JamesMTitus
More Human
Social Friend @JamesMTitus
Knowledge Lajello
More Human
Silent Influencer @Al_AGW
Social Friend @JamesMTitus
Knowledge Lajello
Can Bots do Astroturfing?
Can Bots do Astroturfing?
•  Reach
•  Intention
•  Ease of Action
Can Bots do Astroturfing?
INTENTION / INFLUENCE
•  Opinion leaders (Katz 1955)
•  Influentials (Merton 1968)
•  Mavens & co...
Can Bots do Astroturfing?
To create
Intention
Is not easy
•  Reach
•  Intention
•  Ease of Action
Can Bots do Astroturfing?
Readiness
Multiple Sources
Topic Dependence
•  Reach
•  Intention
•  Ease of Action
To create
Int...
But it is NOT impossible
But it is NOT impossible
Arjun	
  Mukherjee	
  et.al.	
  	
  
Influence the News
50%
Influence the News
50%
55%
Influence the News
0	
   20.000	
   40.000	
   60.000	
   80.000	
   100.000	
   120.000
N	
  Korean	
  leader	
  Kim	
  Jo...
Truth about the Truth
Thus how to spot them?
INDIVIDUAL
•  Not ‘loud’
•  Might be human
•  Missing trainings data
Thus how to spot them?
GROUP
•  Similarity of group
•  Description, Focus,
Tweets…
INDIVIDUAL
•  Not ‘loud’
•  Might be hu...
Outlook
•  Arm’s length Race
•  Verification
•  Avoid Training Spammer
•  The New Gatekeeper
Thanks
MiningData.Biz
LutzFinger.com SoumitraDutta.com
Who is fake  discover astroturfing or attempts of fake influence  presentation
Who is fake  discover astroturfing or attempts of fake influence  presentation
Who is fake  discover astroturfing or attempts of fake influence  presentation
Who is fake  discover astroturfing or attempts of fake influence  presentation
Who is fake  discover astroturfing or attempts of fake influence  presentation
Upcoming SlideShare
Loading in …5
×

Who is fake discover astroturfing or attempts of fake influence presentation

360 views

Published on

Who is Fake? Discover Astroturfing or Attempts of Fake Influence!

Published in: Technology, News & Politics
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
360
On SlideShare
0
From Embeds
0
Number of Embeds
15
Actions
Shares
0
Downloads
8
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Who is fake discover astroturfing or attempts of fake influence presentation

  1. 1. Who is FAKE? Discover Astroturfing or Attempts of Fake Influence! Lutz Finger Soumitra DuttaMiningData.biz
  2. 2. An Army of Bots $ 2.76 million
  3. 3. The Social Net
  4. 4. The Social Net
  5. 5. The Analytics
  6. 6. The Problem
  7. 7. ANY measurement (if useful) Will be ATTACKED.
  8. 8. Mass Movement •  Reach •  Intention •  Ease of Action
  9. 9. Mass Movement •  Reach •  Intention •  Ease of Action
  10. 10. Mass Movement Astroturf •  Reach •  Intention •  Ease of Action
  11. 11. Bots 1.0 – Spammer going Social @you malware.com @you-as-well malware.com D fresh-contact malware.com
  12. 12. Bots 1.0 – Spammer going Social 10.000 messages in 4 month @PeaceKaren_25   Jacob  Ratkiewicz  et.  al    -­‐  2011  
  13. 13. 10.000 messages in 4 month Bots 1.0 – Spammer going Social @PeaceKaren_25   Jacob  Ratkiewicz  et.  al    -­‐  2011  
  14. 14. Look for Un-Normal Training: •  @spam •  Manual classification •  Honey Pots
  15. 15. Look for Un-Normal Differences: •  Time: Regular or Bursty •  Heavy Hashtag Usage •  Blacklisted URL •  Spam Words •  Few Friends Grier  et  al.  2010        (Berkeley)   Training: •  @spam •  Manual classification •  Honey Pots
  16. 16. Thus Social Networks acted….
  17. 17. Thus Social Networks acted…. 2013 7% SPAM 20122011 20% detection
  18. 18. Detected but still Dangerous SMEAR Campaigns Denial of INFORMATION
  19. 19. Bot 2.0 – Social Bots Analytics ToolsConversation Bots
  20. 20. More Human Social Friend @JamesMTitus
  21. 21. More Human Social Friend @JamesMTitus
  22. 22. More Human Social Friend @JamesMTitus Knowledge Lajello
  23. 23. More Human Silent Influencer @Al_AGW Social Friend @JamesMTitus Knowledge Lajello
  24. 24. Can Bots do Astroturfing?
  25. 25. Can Bots do Astroturfing? •  Reach •  Intention •  Ease of Action
  26. 26. Can Bots do Astroturfing? INTENTION / INFLUENCE •  Opinion leaders (Katz 1955) •  Influentials (Merton 1968) •  Mavens & connectors (Gladwell 2000) •  Reach •  Intention •  Ease of Action
  27. 27. Can Bots do Astroturfing? To create Intention Is not easy •  Reach •  Intention •  Ease of Action
  28. 28. Can Bots do Astroturfing? Readiness Multiple Sources Topic Dependence •  Reach •  Intention •  Ease of Action To create Intention Is not easy
  29. 29. But it is NOT impossible
  30. 30. But it is NOT impossible Arjun  Mukherjee  et.al.    
  31. 31. Influence the News 50%
  32. 32. Influence the News 50% 55%
  33. 33. Influence the News 0   20.000   40.000   60.000   80.000   100.000   120.000 N  Korean  leader  Kim  Jong-­‐il  dies   AnMmaNer  atom  trapped  for  first  Mme,  say  scienMsts     Neutrinos  beat  light  speed  again   Earth-­‐like  planet  found  in  the  "habitable  zone"   No  rhinos  remain  in  West  Africa   Eurozone  debt  web:  Who  owes  what  to  whom?   Writer  Christopher  Hitchens  dies   BBC  apology  for  Clarkson  comments   'Witch's  coNage'  found  in  Pendle   In  pictures:  ApocalypMc  Manchester   Social  Media  Index   Clicking   Sharing   CommenMng   50% on  the  courtesy  of       55%
  34. 34. Truth about the Truth
  35. 35. Thus how to spot them? INDIVIDUAL •  Not ‘loud’ •  Might be human •  Missing trainings data
  36. 36. Thus how to spot them? GROUP •  Similarity of group •  Description, Focus, Tweets… INDIVIDUAL •  Not ‘loud’ •  Might be human •  Missing trainings data
  37. 37. Outlook •  Arm’s length Race •  Verification •  Avoid Training Spammer •  The New Gatekeeper
  38. 38. Thanks MiningData.Biz LutzFinger.com SoumitraDutta.com

×