Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup.

13
Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup

Transcript of Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup.

Page 1: Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup.

Spam? No, thanks!

Panos Ipeirotis – New York University

“Crowdsourcing Work” Meetup

Page 2: Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup.

“A Computer Scientist in a Business

School”

http://behind-the-enemy-lines.blogspot

.com/

Email: [email protected]

“A Computer Scientist in a Business

School”

http://behind-the-enemy-lines.blogspot

.com/

Email: [email protected]

Panos Ipeirotis - Introduction

New York University, Stern School of Business

Page 3: Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup.

Example: Build an Adult Web Site Classifier

Need a large number of hand-labeled sites Get people to look at sites and classify them

as:G (general), PG (parental guidance), R (restricted), X

(porn)

Cost/Speed Statistics Undergrad intern: 200 websites/hr, cost:

$15/hr MTurk: 2500 websites/hr, cost: $12/hr

Cost/Speed Statistics Undergrad intern: 200 websites/hr, cost:

$15/hr MTurk: 2500 websites/hr, cost: $12/hr

Page 4: Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup.

Bad news: Spammers!

Worker ATAMRO447HWJQ

labeled X (porn) sites as G (general

audience)

Worker ATAMRO447HWJQ

labeled X (porn) sites as G (general

audience)

Page 5: Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup.

Improve Data Quality through Repeated Labeling

Get multiple, redundant labels using multiple workers Pick the correct label based on majority vote

Probability of correctness increases with number of workers

Probability of correctness increases with quality of workers

1 worker

70%

correct

1 worker

70%

correct

11 workers

93%

correct

11 workers

93%

correct

Page 6: Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup.

11-vote Statistics MTurk: 227 websites/hr, cost: $12/hr Undergrad: 200 websites/hr, cost:

$15/hr

11-vote Statistics MTurk: 227 websites/hr, cost: $12/hr Undergrad: 200 websites/hr, cost:

$15/hr

Single Vote Statistics MTurk: 2500 websites/hr, cost: $12/hr Undergrad: 200 websites/hr, cost:

$15/hr

Single Vote Statistics MTurk: 2500 websites/hr, cost: $12/hr Undergrad: 200 websites/hr, cost:

$15/hr

But Majority Voting is Expensive

Page 7: Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup.

Using redundant votes, we can infer worker quality

Look at our spammer friend ATAMRO447HWJQ together with other 9 workers

Our “friend” ATAMRO447HWJQ mainly marked sites as G.Obviously a spammer…

We can compute error rates for each worker

Error rates for ATAMRO447HWJQ P[X → X]=9.847% P[X → G]=90.153% P[G → X]=0.053% P[G → G]=99.947%

Page 8: Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup.

Rejecting spammers and BenefitsRandom answers error rate = 50%

Average error rate for ATAMRO447HWJQ: 45.2% P[X → X]=9.847% P[X → G]=90.153% P[G → X]=0.053% P[G → G]=99.947%

Action: REJECT and BLOCK

Results: Over time you block all spammers Spammers learn to avoid your HITS You can decrease redundancy, as quality of workers is higher

Page 9: Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup.

After rejecting spammers, quality goes up Spam keeps quality down Without spam, workers are of higher quality Need less redundancy for same quality Same quality of results for lower cost

With spam

1 worker

70%

correct

With spam

1 worker

70%

correct

With spam

11 workers

93%

correct

With spam

11 workers

93%

correct

Without

spam

1 worker

80% correct

Without

spam

1 worker

80% correct

Without

spam

5 workers

94% correct

Without

spam

5 workers

94% correct

Page 10: Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup.

Correcting biases

Classifying sites as G, PG, R, X Sometimes workers are careful but biased

Classifies G → P and P → R Average error rate for ATLJIK76YH1TF: 45.0%

Is ATLJIK76YH1TF a spammer?Is ATLJIK76YH1TF a spammer?

Error Rates for Worker: ATLJIK76YH1TF

P[G → G]=20.0% P[G → P]=80.0%P[G → R]=0.0% P[G → X]=0.0%P[P → G]=0.0% P[P → P]=0.0% P[P → R]=100.0% P[P → X]=0.0%P[R → G]=0.0% P[R → P]=0.0% P[R → R]=100.0% P[R → X]=0.0%P[X → G]=0.0% P[X → P]=0.0% P[X → R]=0.0% P[X → X]=100.0%

Error Rates for Worker: ATLJIK76YH1TF

P[G → G]=20.0% P[G → P]=80.0%P[G → R]=0.0% P[G → X]=0.0%P[P → G]=0.0% P[P → P]=0.0% P[P → R]=100.0% P[P → X]=0.0%P[R → G]=0.0% P[R → P]=0.0% P[R → R]=100.0% P[R → X]=0.0%P[X → G]=0.0% P[X → P]=0.0% P[X → R]=0.0% P[X → X]=100.0%

Page 11: Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup.

Correcting biases

For ATLJIK76YH1TF, we simply need to compute the “non-recoverable” error-rate (technical details omitted)

Non-recoverable error-rate for ATLJIK76YH1TF: 9%

Technical hint: The “condition number” of the matrix [how easy is to invert the matrix] is a good indicator of spamminess

Error Rates for Worker: ATLJIK76YH1TF

P[G → G]=20.0% P[G → P]=80.0%P[G → R]=0.0% P[G → X]=0.0%P[P → G]=0.0% P[P → P]=0.0% P[P → R]=100.0% P[P → X]=0.0%P[R → G]=0.0% P[R → P]=0.0% P[R → R]=100.0% P[R → X]=0.0%P[X → G]=0.0% P[X → P]=0.0% P[X → R]=0.0% P[X → X]=100.0%

Error Rates for Worker: ATLJIK76YH1TF

P[G → G]=20.0% P[G → P]=80.0%P[G → R]=0.0% P[G → X]=0.0%P[P → G]=0.0% P[P → P]=0.0% P[P → R]=100.0% P[P → X]=0.0%P[R → G]=0.0% P[R → P]=0.0% P[R → R]=100.0% P[R → X]=0.0%P[X → G]=0.0% P[X → P]=0.0% P[X → R]=0.0% P[X → X]=100.0%

Page 12: Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup.

Too much theory?

Open source implementation available at:http://code.google.com/p/get-another-label/

Input: – Labels from Mechanical Turk– Cost of incorrect labelings (e.g., XG costlier than GX)

Output: – Corrected labels– Worker error rates– Ranking of workers according to their quality

Alpha version, more improvements to come! Suggestions and collaborations welcomed!

Page 13: Spam? No, thanks! Panos Ipeirotis – New York University “Crowdsourcing Work” Meetup.

Thank you!

Questions?

“A Computer Scientist in a Business School”

http://behind-the-enemy-lines.blogspot.com

/

Email: [email protected]

“A Computer Scientist in a Business School”

http://behind-the-enemy-lines.blogspot.com

/

Email: [email protected]