Managing Crowdsourced Human Computation: A Tutorial

Managing Crowdsourced Human ComputationManaging Crowdsourced Human Computation Panos Ipeirotis, New York University Slides from the WWW2011 tutorial, 29 March 2011 Slides from the WWW2011 tutorial 29 March 2011

Outline• Introduction: Human computation and crowdsourcing• Managing quality for simple tasks• Complex tasks using workflows• Task optimization• Incentivizing the crowd Incentivizing the crowd• Market design• Behavioral aspects and cognitive biases Behavioral aspects and cognitive biases• Game design• Case studies d

Human Computation, Round 1 • Humans were the first “computers,” used for computers, used for math computations Grier, When computers were human, 2005 Grier, IEEE Annals 1998

Human Computation, Round 1 • Humans were the first “computers,” used for math computations • Organized computation: – Clairaut, astronomy, 1758: y Computed the Halley’s comet orbit (three‐body problem) dividing the problem) dividing the labor of numeric computations across 3 astronomers Grier, When computers were human, 2005 Grier, IEEE Annals 1998

Human Computation, Round 1 • Organized computation: – Maskelyne, astronomical almanac with moon positions, used for p , navigation, 1760. Quality assurance by doing calculations twice and compared by third verifier. verifier – De Prony, 1794, hires hairdressers (unemployed after French (unemployed after French revolution; knew only addition and subtraction) to create logarithmic and trigonometric tables. He managed the process tables He managed the process by splitting the work into very detailed workflows. (Hairdressers better than mathematicians in arithmetic!) Grier, When computers were human, 2005 Grier, IEEE Annals 1998

Human Computation, Round 1 • Organized computation: – Clairaut, astronomy, 1758 – Maskelyne, 1760 – De Prony, log/trig tables, 1794 – Galton, biology, 1893 – Pearson, biology, 1899 – … – Cowles, stock market, 1929 – Math Tables Project, unskilled Math Tables Project unskilled labor, 1938 Grier, When computers were human, 2005 Grier, IEEE Annals 1998

Human Computation, Round 1 • Patterns emerging Patterns emerging – Division of labor – Mass production Mass production – Professional managers • Then we got the “automatic computers”

Human Computation, Round 2 • Now we need humans again for the “AI‐complete” i f th “AI l t ” tasks – Tag images [ESP Game: voh Ahn and Dabbish 2004, ImageNet] – Determine if page relevant Determine if page relevant [Alonso et al., 2011] – Determine song genre – Check page for offensive Check page for offensive content –… ImageNet: http://www.image‐net.org/about‐publication

Focus of the tutorial Focus of the tutorial Examine cases where humans interact with Examine cases where humans interact withcomputers in order to solve a computational problem (usually too hard to be solved by (usually too hard to be solved by computers alone)

Crowdsourcing and human computation• Crowdsourcing: From macro to micro g – Netflix, Innocentive – Quirky, Threadless – oDesk, Guru, eLance, vWorker D k G L W k – Wikipedia et al. – ESP Game, FoldIt, Phylo, … S Ga e, o d t, y o, – Mechanical Turk, CloudCrowd, …• Crowdsourcing greatly facilitates human computation (but they are not equivalent)

Micro‐Crowdsourcing Example: Labeling Images L b li I using the ESP Game Luis von Ahn MacArthur Fellowship "genius grant" • Two player online game • Partners don’t know each other and can’t communicate • Object of the game: type the same word • The only thing in common is an image The only thing in common is an image

PLAYER 1 PLAYER 2 GUESSING: CAR GUESSING: BOY GUESSING: HAT GUESSING: CARGUESSING: KID SUCCESS! YOU AGREE ON CARSUCCESS!YOU AGREE ON CAR

Paid Crowdsourcing: Amazon Mechanical Turk

Demographics of MTurk workers http://bit.ly/mturk‐demographics http://bit.ly/mturk demographicsCountry of residenceCountry of residence• United States: 46.80%• India: 34.00%• Miscellaneous: 19.20%

Demographics of MTurk workers http://bit.ly/mturk‐demographics

Demographics of MTurk workers http://bit.ly/mturk‐demographics http://bit.ly/mturk demographics

Managing quality for simple tasks Managing quality for simple tasks• Quality through redundancy: Combining votes Quality through redundancy: Combining votes – Majority vote – Quality adjusted vote Quality‐adjusted vote – Managing dependencies• Quality through gold data Q lit th h ld d t• Estimating worker quality (Redundancy + Gold)• Joint estimation of worker quality and difficulty• Active data collection Active data collection

Example: Build an “Adult Web Site” Classifier• Need a large number of hand‐labeled sites• G t Get people to look at sites and classify them as: l t l k t it d l if thG (general audience) PG (parental guidance) R (restricted) X (porn)

Example: Build an “Adult Web Site” Classifier • Need a large number of hand‐labeled sites • G t Get people to look at sites and classify them as: l t l k t it d l if th G (general audience) PG (parental guidance) R (restricted) X (porn)Cost/Speed Statistics Undergrad intern: 200 websites/hr, cost: $15/hr

Example: Build an “Adult Web Site” Classifier • Need a large number of hand‐labeled sites • G t Get people to look at sites and classify them as: l t l k t it d l if th G (general audience) 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

Bad news: Spammers! Worker ATAMRO447HWJQlabeled X (porn) sites as G (general audience)

Majority Voting and Label Quality  Ask multiple labelers, keep majority label as “true” label  Quality is probability of being correct 1 p=1.0 0.9 p=0.9 09 0.8 p=0.7 Quality for Majority Vote p=0.8 0.7 p=0.6 p is probability 0.6 of individual labeler of individual labeler p=0.5 p=0 5 M 0.5 being correct 0.4 p=0.4 0.3 0.2 1 3 5 7 9 11 13 Number of labelers Binary classification Binary classification 26Kuncheva et al., PA&A, 2003

What if qualities of workers are different? 3 workers, qualities: p‐d, p, p+d Region where majority better• Majority vote works best when workers have similar quality j p• Otherwise better to just pick the vote of the best worker• …or model worker qualities and combine [coming next]

Combining votes with different quality Clemen and Winkler, 1990

What happens if we have dependencies? Clemen and Winkler, 1985 Positive dependencies decrease the number of effective labelers Positive dependencies decrease the number of effective labelers

What happens if we have dependencies? Yule’s Q Y l ’ Q measure of correlation Kuncheva et al., PA&A, 2003 Positive dependencies decrease the number of effective labelers Positive dependencies decrease the number of effective labelers Negative dependencies can improve results (unlikely both workers to be wrong at the same time)

Vote combination: Meta studies Vote combination: Meta‐studies• Simple averages tend to work well Simple averages tend to work well• C Complex models slightly better but less robust l d l li h l b b l b [Clemen and Winkler, 1999, Ariely et al. 2000]

From aggregate labels to worker quality Look at our spammer friend ATAMRO447HWJQ h ih h 9 k together with other 9 workersAfter aggregation, we compute confusion matrix for each worker After majority vote, confusion matrix for ATAMRO447HWJQ P[G → G]=100% P[G → X]=0% P[X → G]=100% P[X → X]=0%

Algorithm of Dawid & Skene, 1979 Iterative process to estimate worker error rates1. Initialize by aggregating labels for each object (e.g., use majority vote)2. Estimate confusion matrix for workers (using aggregate labels) l b l )3. Estimate aggregate labels (using confusion matrix) • Keep labels for “gold data unchanged gold data”4. Go to Step 2 and iterate until convergence Confusion matrix for ATAMRO447HWJQ Our f i d O friend ATAMRO447HWJQ P[G → G]=99.947% P[G → X]=0.053% marked almost all sites as G. P[X → G]=99.153% P[X → X]=0.847% Seems like a spammer…

And many variations… And many variations…• van der Linden et al, 1997: Item‐Response Theory a de de et a , 99 : te espo se eo y• Uebersax, Biostatistics 1993: Ordered categories• Uebersax, JASA 1993: Ordered categories, with worker Uebersax, JASA 1993: Ordered categories, with worker expertise and bias, item difficulty• Carpenter, 2008: Hierarchical Bayesian versions p , yAnd more recently at NIPS: y• Whitehill et al., 2009: Adding item difficulty• Welinder et al., 2010: Adding worker expertise et al., 2010: Adding worker expertise

Challenge: From Confusion Matrixes to Quality Scores All the algorithms will generate “confusion matrixes” for workers Confusion C f i matrix for ti f ATAMRO447HWJQ P[X → X]=0.847% P[X → G]=99.153% P[G → X]=0.053% P[G → G]=99.947% How to check if a worker is a spammer using the confusion matrix? g (hint: error rate not enough)

Challenge 1: Spammers are lazy and smart! Spammers are lazy and smart!Confusion matrix for spammer Confusion matrix for good worker P[X → X]=0% P[X → G]=100% X] 0% G] 100%  P[X → X]=80% P[X → G]=20% P[G → X]=0% P[G → G]=100%  P[G → X]=20% P[G → G]=80% • Spammers figure out how to fly under the radar… • I In reality, we have 85% G sites and 15% X sites lit h 85% G it d 15% X it • Errors of spammer = 0% * 85% + 100% * 15% = 15% Errors of spammer 0% 85% + 100% 15% 15% • Error rate of good worker = 85% * 20% + 85% * 20% = 20% False negatives: Spam workers pass as legitimate

Challenge 2: Humans are biased! Humans are biased!Error rates for CEO of AdSafe 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% G]=0.0% P[X → P] 0 0% P]=0.0% P[X → R] 0 0% R]=0.0% P[X → X] 100 0% X]=100.0% In reality, we have 85% G sites, 5% P sites, 5% R sites, 5% X sites Errors of spammer (all in G) = 0% * 85% + 100% * 15% = 15% Error rate of biased worker = 80% * 85% + 100% * 5% = 73%False positives: Legitimate workers appear to be spammers

Solution: Reverse errors first, compute f d error rate afterwardsError Rates for CEO of AdSafe P[G → G]=20 0% G]=20.0% P[G → P]=80 0% P]=80.0% P[G → R]=0 0% R]=0.0% P[G → X]=0 0% 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% [ ] • When biased worker says G, it is 100% G • When biased worker says P, it is 100% G Wh bi d k P it i 100% G • When biased worker says R, it is 50% P, 50% R • When biased worker says X, it is 100% X Small ambiguity for “R‐rated” votes but other than that, fine!

Solution: Reverse errors first, compute f d error rate afterwardsError Rates for spammer: ATAMRO447HWJQ P[G → G]=100.0% P[G → P]=0.0% P[G → R]=0.0% P[G → X]=0.0% P[P → G]=100.0% P[P → P]=0.0% P[P → R]=0.0% P[P → X]=0.0% P[R → G]=100.0% P[R → P]=0.0% P[R → R]=0.0% P[R → X]=0.0% P[X → G]=100 0% G]=100.0% P[X → P]=0 0% P]=0.0% P[X → R]=0 0% R]=0.0% P[X → X]=0 0% X]=0.0% • When spammer says G, it is 25% G, 25% P, 25% R, 25% X • When spammer says P, it is 25% G, 25% P, 25% R, 25% X • When spammer says R, it is 25% G, 25% P, 25% R, 25% X • When spammer says X, it is 25% G, 25% P, 25% R, 25% X [note: assume equal priors] The results are highly ambiguous. No information provided! h l h hl b f

Quality Scores• High cost when “soft” labels have probability spread across classes• Low cost when “soft” labels have probability mass concentrated in one class Assigned Label “Soft” Label Cost G <G: 25%, P: 25%, R: 25%, X: 25%> 0.75 G <G: 99%, P: 1%, R: 0%, X: 0%> 0.0198 [***Assume equal misclassification costs] Ipeirotis, Provost, Wang, HCOMP 2010

Quality Score• A spammer is a worker who always assigns labels randomly, regardless of what the true class is. QualityScore = 1 ‐ ExpCost(Worker)/ExpCost(Spammer)• Q li S QualityScore i is useful for the purpose of blocking bad workers and f lf h f bl ki b d k d rewarding good ones• Essentially a multi class cost sensitive AUC metric Essentially a multi‐class, cost‐sensitive AUC metric • AUC = area under the ROC curve

What about Gold testing? Naturally integrated into the latent class model1. (e.g.,1 Initialize by aggregating labels for each object (e g use majority vote)2. Estimate error rates for workers (using aggregate labels)3. Estimate aggregate labels (using error rates, weight worker votes according to quality) • Keep labels for “gold data” unchanged4 Go to Step 2 and iterate until convergence4.

• 3 labels per example • 2 categories, 50/50 Gold Testing Gold Testing • • Quality range: 0.55:0.05:1.0 200 labelers l b l No significant advantage under “good conditions” (balanced datasets, good worker quality)http://bit.ly/gold‐or‐repeatedWang, Ipeirotis, Provost, WCBI 2011

• 5 labels per example • 2 categories, 50/50Gold TestingGold Testing • • Quality range: 0.55:1.0 200 labelers l b l No significant advantage under “good conditions” (balanced datasets, good worker quality)

• 10 labels per example • 2 categories, 50/50Gold TestingGold Testing • • Quality range: 0.55:1.0 200 labelers l b l No significant advantage under “good conditions” (balanced datasets, good worker quality)

• 10 labels per example • 2 categories, 90/10Gold TestingGold Testing • • Quality range: 0.55:0.1.0 l b l 200 labelers Advantage under imbalanced datasets

• 5 labels per example • 2 categories, 50/50Gold TestingGold Testing • • Quality range: 0.55:0.65 200 labelers l b l Advantage with bad worker quality

• 10 labels per example • 2 categories, 90/10Gold Testing?Gold Testing? • • Quality range: 0.55:0.65 200 labelers l b l Significant advantage under “bad conditions” (imbalanced datasets, bad worker quality)

Testing workers Testing workers• An exploration‐exploitation scheme: An exploration exploitation – Explore: Learn about the quality of the workers – Exploit: Label new examples using the quality Exploit: Label new examples using the quality

Testing workers Testing workers• An exploration‐exploitation scheme: An exploration exploitation – Assign gold labels when benefit in learning better quality of worker outweighs the loss for labeling a quality of worker outweighs the loss for labeling a gold (known label) example [Wang et al, WCBI 2011] – Assign an already labeled example (by other Assign an already labeled example (by other workers) and see if it agrees with majority [Donmez et al., KDD 2009] – If worker quality changes over time, assume f k l h accuracy given by HMM and φ(τ) = φ(τ‐1) + Δ [ [Donmez et al., SDM 2010] , ]

Example: Build an “Adult Web Site” Classifier Example: Build an Adult Web Site Get people to look at sites and classify them as: p p y G (general audience) PG (parental guidance) R (restricted) X (porn)But we are not going to label the whole Internet…ExpensiveSlow

Integrating with Machine LearningIntegrating with Machine Learning• Crowdsourcing is cheap but not free Crowdsourcing is cheap but not free – Cannot scale to web without help• Solution: Build automatic classification models using crowdsourced data

Simple solution• Humans label training data Humans label training data• Use training data to build model Data from existing crowdsourced answersNNew C Case Automatic Model Automatic (through machine learning) Answer

Quality and Classification Performance Noisy labels lead to degraded task performance Labeling quality increases  classification quality increases Labeling quality increases classification quality increases Quality = 100% 100 Quality 80% Quality = 80% 90 80 AUC Quality = 60% 70 60 Quality = 50% 50 40 100 120 140 160 180 200 220 240 260 280 300 1 20 40 60 80 Number of examples ("Mushroom" data set) Single‐labeler quality (p (probability of assigning y g 54 g correctly a binary label)http://bit.ly/gold‐or‐repeatedSheng, Provost, Ipeirotis, KDD 2008

Tradeoffs for Machine Learning Models • Get more data  Improve model accuracy • Improve data quality  Improve classification p q y p Data Quality = 100% 100 Data Quality = 80% Q li 80% 90 80 uracy 70 Data Quality = 60% Q y Accu 60 50 Data Quality = 50% 40 0 0 0 0 0 0 0 0 0 0 0 1 20 40 60 80 10 12 14 16 18 20 22 24 26 28 30 Number of examples (Mushroom)55

Tradeoffs for Machine Learning Models • Get more data: Active Learning, select which unlabeled example to label [Settles, http://active‐learning.net/] unlabeled example to label [S ttl htt // ti l i t/] • Impro e data q alit Improve data quality: Repeated Labeling, label again an already labeled example [Sheng et al. 2008, Ipeirotis et al, 2010] example [Sheng et al 2008 Ipeirotis et al 2010]56

Scaling Crowdsourcing: Iterative training• Use model when confident, humans otherwise• Retrain with new human input → improve Retrain with new human input → improve model → reduce need for humans Automatic AnswerNew Case Automatic Model (through machine learning) Data from existing Get human(s) to answer crowdsourced answers

Rule of Thumb Results Rule of Thumb Results• With high quality labelers (80% and above): One g q y ( ) worker per case (more data better)• With low quality labelers (~60%) Multiple workers per case (to improve quality) Multiple workers per case (to improve quality)[Sheng et al KDD 2008; Kumar and Lease CSDM 2011] et al, KDD 2008; Kumar and Lease, CSDM 2011] 58

Dawid & Skene meets a Classifier & Skene meets a Classifier• [Raykar et al. JMLR 2010]: Use the et al. JMLR 2010]: Use the Dawid&Skene scheme but add a classifier as an additional worker an additional worker• Classifier in each iteration learns from the consensus labeling 59

Selective Repeated Labeling Selective Repeated‐Labeling• We do not need to label everything same number of times• Key observation: we have additional information to guide selection of data for repeated labeling p g the current multiset of labels • Example: {+ ‐ + ‐ ‐ +} vs {+ + + + + +} Example: {+,‐,+,‐,‐,+} vs. {+,+,+,+,+,+} 60

Label Uncertainty: Focus on uncertainty• If we know worker qualities, we can estimate log‐odds for each q , g example:• Assign labels first to examples that are most uncertain (log‐ odds close to 0 for binary case)

+ + ‐ ‐‐ ‐ + + + + + + ‐ ‐‐ ‐ ‐ + ‐ ‐ ‐Model Uncertainty (MU) + + + + + + ++ + +‐ ‐ ‐ ‐‐ ‐ ‐ + + + + ‐‐‐‐ ‐‐‐‐ ‐‐ ‐‐• Learning models of the data provides an alternative source of information about label certainty• M d l Model uncertainty: get more labels for instances t i t t l b l f i t Examples that cause model uncertainty• Intuition? Models – for modeling: why improve training data quality if “Self‐healing” process model already is certain there? [Brodley et al, JAIR 1999] [Ipeirotis et al NYU 2010] et al, NYU 2010] – for data quality, low‐certainty “regions” may be due to incorrect labeling of corresponding instances 62

Adult content classification Round RobinSelective labeling 63

Too much theory? Too much theory? Open source implementation available at: p p http://code.google.com/p/get‐another‐label/• Input: – Labels from Mechanical Turk – Cost of incorrect labelings (e.g., XG costlier than GX)• Output: Output: – Corrected labels – Worker error rates – Ranking of workers according to their quality

Learning from imperfect data Learning from imperfect data 100• With inherently noisy With inherently noisy 90 80 Accuracy data, good to have 70 60 learning algorithms that learning algorithms that 50 are robust to noise. 40 0 0 0 0 0 0 0 0 0 0 0 1 20 40 60 80 18 20 22 24 26 28 30 10 12 14 16 Number of examples (Mushroom)• Or use techniques designed to handle d i d h dl explicitly noisy data [Lugosi 1992; Smyth, 1995, 1996]

How to handle free‐form answers?• Q: “My task does not have discrete answers….”• A: Break into two HITs: – “C t ” HIT “Create” – “Vote” HIT Creation HIT Voting HIT: (e.g. find a URL about a topic) Correct or not?• Vote HIT controls quality of Creation HIT Vote HIT controls quality of Creation HIT• Redundancy controls quality of Voting HIT• Catch: If “creation” very good, in voting workers just vote “yes” – Solution: Add some random noise (e.g. add typos) Example: Collect URLs

But my free‐form is Describe this j g gnot just right or wrong…• “Create” HIT• “Improve” HIT p• “Compare” HIT Creation HIT ( g (e.g. describe the image) g ) Improve HIT Compare HIT (voting)(e.g. improve description) Which is better? TurkIt toolkit [Little et al., UIST 2010]: http://groups.csail.mit.edu/uid/turkit/

version 1: A parial view of a pocket calculator together with some coins and a pen.version 2:version 2: A view of personal items a calculator, and some gold and copper coins, and a round tip pen, these are all pocket and wallet sized item used for business, writting, calculating prices or solving math problems and purchasing items.version 3: A close‐up photograph of the following items: A CASIO multi‐function calculator. A ball point pen, uncapped. Various coins, apparently European, both copper and gold. Seems to be a theme illustration for a brochure or document b h ll f b h d cover treating finance, probably personal finance.version 4: …Various British coins; two of £1 value, three of 20p value and one of 1p value. … and one of 1p valueversion 8: “A close‐up photograph of the following items: A CASIO multi‐function, solar powered scientific calculator. A blue ball point pen with a blue rubber grip and the tip extended. Six British coins; two of £1 value, three of 20p value and one of 1p value. Seems to be a theme illustration for a brochure or document cover treating finance ‐ probably personal finance."

Independence or Not? Independence or Not? • Building iteratively (lack of independent) allows better outcomes for image description task… • In the FoldIt game workers built on each other’s results In the FoldIt game, workers built on each other s results[Little et al, HCOMP 2010]

Independence or Not? • But lack of independence But lack of independence may cause high dependence on starting conditions and create conditions and create groupthink p • …but also prevents disasters[Little et al, HCOMP 2010]

Independence or Not? Collective Problem Solving • Exploration / exploitation tradeoff – Can accelerate learning, by sharing good solutions – But can lead to premature convergence on suboptimal solution[Mason and Watts, submitted to Science, 2011]

Individual search strategy affects group successIndividual search strategy affects group success • More players copying More players copying each other (i.e., fewer exploring) in current round  Lower probability of finding peak on next roundd

The role of Communication NetworksThe role of Communication Networks• Examine various “neighbor” structures g (who talks to whom about the oil levels)

Network structure affects individual search strategy• Higher clustering  Higher probability of neighbors guessing in neighbors guessing in identical location• More neighbors guessing in identical location  Higher probability of copying

Diffusion of Best SolutionDiffusion of Best Solution

Individual search strategy affects group successIndividual search strategy affects group success• No significant No significant differences in % of games in which peak was found• Network affects willingness to explore

Network structure affects group successNetwork structure affects group success

TurKontrol: Decision Theoretic ModelingTurKontrol: Decision‐Theoretic Modeling• Optimizing workflow execution using decision Optimizing workflow execution using decision‐ theoretic approaches [Dai et al, AAAI 2010; Kern et al. 2010]• Si ifi Significant work in control theory [Montgomery, 2007] ki l h

http://www.workflowpatterns.com Common Workflow Patterns Common Workflow PatternsBasic Control FlowBasic Control Flow • Iteration• Sequence • Arbitrary Cycles (goto)• Parallel Split • Structured Loop (for, while, repeat) while, repeat)• Synchronization • Recursion• Exclusive Choice• Simple Merge

Soylent• Word processor with crowd embedded [Bernstein et al, UIST 2010]• “Proofread paper”: Ask workers to proofread each paragraph –LLazy Turker: Fixes the minimum possible (e.g., single typo) T k Fi th i i ibl ( i l t ) – Eager Beaver: Fixes way beyond the necessary but adds extra errors (e.g., inline suggestions on writing style)• Find‐Fix‐Verify pattern – Separate Find and Fix, does not allow Lazy Turker – Separate Fix‐Verify ensured quality

Find “Identify at least one area t at can that ca be s o te ed shortened without changing the meaning of the paragraph.” Independent agreement to identify patchesFix “Edit the highlighted g g section to shorten its length without changing Soylent, a prototype... the meaning of the paragraph.” Randomize order of suggestionsVerify “Choose at least one rewrite that has style errors, and at least one rewrite that changes the meaning of the sentence.”

Crowd‐created Workflows: CrowdForge• Map‐Reduce framework for crowds [Kittur et al, CHI 2011] – Identify sights worth checking out ( (one tip per worker)) • Vote and rank – Brief tips for each monument ( (one tip p worker) p per ) • Vote and rank – Aggregate tips in meaningful summary • It t to improve… Iterate t iMy Boss is a Robot (mybossisarobot.com), Nikki Kittur (CMU) + Jim Giles (New Scientist)

Crowd‐created Workflows: TurkoMatic• Crowd creates workflows• Turkomatic [Kalkani et al, CHI 2011]: 1. Ask workers to decompose task into steps (Map) 2. Can step be completed within 10 minutes? 1. Yes: solve it. 2. 2 No: decompose further (recursion) 3. Given all partial solutions, solve big problem (Reduce)

Crowdsourcing Patterns Crowdsourcing Patterns• Generate / Create i d• Find Creation C ti• Improve / Edit / Fix• Vote for accept‐reject Q Quality y• Vote up, vote down, to generate rank Control• Vote for best / select top‐k• Split task• Aggregate Flow Control Flow Control

Defining Task Parameters Defining Task ParametersThree main goals:Three main goals:• Minimize Cost (cheap) i i i C ( h )• Maximize Quality (good)• Minimize Completion Time (fast)

Effect of Payment: Quality Effect of Payment: Quality• Cost does not affect quality [Mason and Watts, 2009, AdSafe]• Similar results for bigger tasks [Ariely et al, 2009] 0.45 0.40 0.35 0.30 Error Rate 0.25 2cents 5cents 0.20 10cents 0.15 0 0.10 0.05 0.00 0 10 20 30 Number of Labelers

Effect of Payment: #Tasks Effect of Payment: #Tasks• Payment incentives increase speed, though [Mason and Watts, 2009]

Predicting Completion Time• Model timing of individual task [Yan, Kumar, Ganesan, 2010] – Assume rate of task completion λ – Exponential distribution for single task g – Erlang distribution for sequential tasks – On the fly estimation of λ for On‐the‐fly estimation of λ for parallel• Optimize using early acceptance/termination – Sequential experiment setting – Stop early if confident Stop early if confident

Prediction Completion Time Prediction Completion Time• For Freebase, workers use log‐normal time to complete a task [Kochhar et al, HCOMP 2010]

Predicting Completion Time• Exponential assumption usually not realistic• H il d di ib i Heavy‐tailed distribution [Ipeirotis, XRDS 2010]

Effect of #HITs: Monotonic, but sublinear h(t) = 0.998^#HITs• 10 HITs  2% slower than 1 HIT• 100 HITs  19% slower than 1 HIT • 1000 HITs  87% slower than 1 HIT 1000 HITs  87% slower than 1 HIT or, 1 group of 1000  7 times faster than 1000 sequential groups of 1 [Wang et al, CSDM 2011]

HIT Topicstopic 1 : cw castingwords podcast transcribe english mp3 edit confirm snippet gradetopic 2: d i 2 data collection search image entry listings website review survey opinion ll i h i li i bi i i itopic 3: categorization product video page smartsheet web comment website opiniontopic 4: easy quick survey money research fast simple form answers linktopic 5: question answer nanonano dinkle article write writing review blog articlestopic 6: writing answer article question opinion short advice editing rewriting paultopic 7: transcribe transcription improve retranscribe edit answerly voicemail answer [Wang et al, CSDM 2011]

Effect of Topic: The CastingWords Effect topic 1 : cw castingwords podcast transcribe english mp3 edit confirm snippet grade topic 2: data collection search image entry listings website review survey opinion topic 3: categorization product video page smartsheet web comment website opinion topic 4: easy quick survey money research fast simple form answers link topic 5: question answer nanonano dinkle article write writing review blog articles p q g g topic 6: writing answer article question opinion short advice editing rewriting paul topic 7: transcribe transcription improve retranscribe edit answerly voicemail query question answer [Wang et al, CSDM 2011]

Effect of Topic: Surveys=fast (even with redundancy!) topic 1 : cw castingwords podcast transcribe english mp3 edit confirm snippet grade topic 2: data collection search image entry listings website review survey opinion topic 3: categorization product video page smartsheet web comment website opinion topic 4: easy quick survey money research fast simple form answers link topic 5: question answer nanonano dinkle article write writing review blog articles p q g g topic 6: writing answer article question opinion short advice editing rewriting paul topic 7: transcribe transcription improve retranscribe edit answerly voicemail query question answer [Wang et al, CSDM 2011]

Effect of Topic: Writing takes time topic 1 : cw castingwords podcast transcribe english mp3 edit confirm snippet grade topic 2: data collection search image entry listings website review survey opinion topic 3: categorization product video page smartsheet web comment website opinion topic 4: easy quick survey money research fast simple form answers link topic 5: question answer nanonano dinkle article write writing review blog articles p q g g topic 6: writing answer article question opinion short advice editing rewriting paul topic 7: transcribe transcription improve retranscribe edit answerly voicemail query question answer [Wang et al, CSDM 2011]

Optimizing Completion Time Optimizing Completion Time• Workers pick tasks that have large number of Workers pick tasks that have large number of HITs or are recent [Chilton et al., HCOMP 2010]• VizWiz optimizations [[Bingham, UIST 2011] : optimizations i h S 20 ] – Posts HITs continuously (to be recent) – Mk bi HIT Mkes big HIT groups (to be large) ( b l ) – HITs are “external HITs” (i.e., IFRAME hosted) – HITs populated when the worker accepts them

Optimizing Completion Time Optimizing Completion Time• Completion rate varies with Completion rate varies with time of day, depending on the audience location (India the audience location (India vs US vs Middle East)• Quality tends to remain the same, independent of same independent of completion time [Huang et al., HCOMP 2010] [Huang et al., HCOMP 2010]

Other Optimizations Other Optimizations• Qurk [Markus et al., CIDR 2011] and CrowdDB [Franklin et al., SIGMOD 2011]: [ ] Treat humans as uncertain UDFs + apply relational optimization, plus the “GoodEnough” and “StopAfter” operator. operator• CrowdFlow [Quinn et al ]: Integrate crowd with machine [Quinn et al.]: Integrate crowd with machine learning to reach balance of speed, quality, cost• Ask humans for directions in a graph: [Parameswaran et al., VLDB 2011]. See also [Kleinberg, Nature 2000; Mitzenmacher, XRDS 2010; Deng, ECCV 2010] Mitzenmacher XRDS 2010; Deng ECCV 2010]

Incentives• Monetary• Self‐serving• Altruistic l i i

Incentives: Money Incentives: Money• Money does not improve quality but (generally) Money does not improve quality but (generally) increase participation [Ariely, 2009; Mason & Watts, 2009]• But workers may be “target earners” (stop after reaching their daily goal) [Horton & Chilton, 2010 for MTurk; hi h i d il l) Camerer et al. 1997, Farber 2008, for taxi drivers; Fehr and Goette 2007]

Incentives: Money and Trouble Incentives: Money and Trouble• Careful: Paying a little often worse than paying y g p y g nothing! – “Pay enough or not at all” [Gneezy et al, 2000] – Small pay now locks future pay Small pay now locks future pay – Payment replaces internal motivation (paying kids to collect donations decreased enthusiasm; spam classification; “thanks for dinner, here is $100 ) dinner here is $100”) – Lesson: Be the Tom Sawyer (“how I like painting the fence”), not the scrooge‐y boss…• Paying a lot is a counter‐incentive: – People focus on the reward and not on the task People focus on the reward and not on the task – On MTurk spammers routinely attack highly‐paying tasks

Incentives: Leaderboards Incentives: Leaderboards• Leaderboards (“top participants”) frequent ( top participants ) frequent motivator – Should motivate correct behavior not just Should motivate correct behavior, not just measurable behavior – Newcomers should have hope of reaching top Newcomers should have hope of reaching top – Whatever is measured, workers will optimize for this (e.g., Orkut country leaderboard; complaints for quality score drops) this (e.g., Orkut country leaderboard; complaints for quality score drops) – Design guideline: Christmas‐tree dashboard (Green / Red lights only) [Farmer and Glass, 2010]

Incentives: Purpose of Work Incentives: Purpose of Work• Contrafreeloading: Rats and animals prefer to Contrafreeloading: Rats and animals prefer to “earn” their food• Destroying work after production demotivates workers. [Ariely et al, 2008] k [A i l l 2008]• Showing result of “completed task” improves satisfaction

Incentives: Purpose of Work Incentives: Purpose of Work• Workers enjoy learning new skills (oft cited reason for j y g ( Mturk participation)• Design tasks to be educational – DuoLingo: Translate while learning new language [von Ahn et al, duolingo.com] – Galaxy Zoo, Clickworkers: Classify astronomical objects [Raddick et al, 2010; http://en.wikipedia.org/wiki/Clickworkers] – Citizen Science: Learn about biology gy [http://www.birds.cornell.edu/citsci/] – National Geographic “Field Expedition: Mongolia”, tag potential archeological sites, learn about archeology potential archeological sites, learn about archeology

Incentives: Credit and ParticipationIncentives: Credit and Participation• Public credit contributes to sense of Public credit contributes to sense of participation• Credit also a form of reputation Credit also a form of reputation• (Anonymity of MTurk‐like settings discourage this factor)

Incentive: Altruism Incentive: Altruism• Contributing back (tit for tat): Early reviewers Contributing back (tit for tat): Early reviewers writing reviews because read other useful review• Effect amplified in social networks: “If all my Effect amplified in social networks: If all my friends do it…” or “Since all my friends will see this…”• Contributing to shared goal Contributing to shared goal

Incentives: Altruism and Purpose Incentives: Altruism and Purpose• On MTurk [Chandler and Kapelner 2010] On MTurk [Chandler and Kapelner, 2010] – Americans [older, more leisure‐driven] work harder for meaningful work harder for “meaningful work” – Indians [more income‐driven] were not affected – Quality unchanged for both groups Quality unchanged for both groups

Incentives: Fair share Incentives: Fair share• Anecdote: Same HIT (spam classification) Anecdote: Same HIT (spam classification) – Case 1: Requester doing as side‐project, to “clean the market would be out‐of‐pocket expense no the market”, would be out of pocket expense, no pay to workers – Case 2: Requester researcher at university, spam Case 2: Requester researcher at university, spam classification now a university research project, payment to workers What setting worked best?

Incentives: FUN! Incentives: FUN!• Game‐ify the task (design details later)• Examples p – ESP Game: Given an image, type the same word (generated image descriptions) – Phylo: aligned color blocks (used for genome Phylo: aligned color blocks (used for genome alignment) – FoldIt: fold structures to optimize energy (protein folding) (protein folding)• Fun factors [Malone 1980, 1982]: – ti d timed response – score keeping – player skill level – high‐score lists – and randomness

Market Design Organizes the CrowdMarket Design Organizes the Crowd• Reputation Mechanisms eputat o ec a s s – Seller‐side: Ensure worker quality – Buy‐side: Ensure employee trustworthiness• Task organization for task discovery (worker finds employer/task) / )• Worker expertise recording for task assignment (employer/task finds worker)

Lack of Reputation and Market for LemonsLack of Reputation and Market for Lemons• “When quality of sold good is uncertain and hidden before transaction, prize goes to value of lowest valued good transaction, prize goes to value of lowest valued good” [Akerlof, 1970; Nobel prize winner]Market evolution steps:1. Employers pays $10 to good worker, $0.1 to bad worker2. 50% good workers, 50% bad; indistinguishable from each other2 50% d k 50% b d d h bl f h h3. Employer offers price in the middle: $54. Some good workers leave the market (pay too low)4 Some good workers leave the market (pay too low)5. Employer revised prices downwards as % of bad increased6. More good workers leave the market… death spiral g p http://en.wikipedia.org/wiki/The_Market_for_Lemons

Lack of Reputation and Market for Lemons Lack of Reputation and Market for Lemons• Market for lemons also on the employer side: – Workers distrust (good) newcomer employers: Charge risk premium, or work only for little bit. Good newcomers get disappointed – Bad newcomers have no downside (will not pay), continue to offer work. work – Market floods with bad employers• TurkOpticon, external reputation system• “Mechanical Turk: Now with 40.92% spam” http://bit.ly/ew6vg4 • Greshams Law: the bad drives out the good• No‐trade equilibrium: no good employer offers work in a No trade equilibrium: no good employer offers work in a market with bad workers, no good worker wants to work for bad employers…• In reality, we need to take into consideration that this is a In reality, we need to take into consideration that this is a repeated game (but participation follows a heavy tail…) http://en.wikipedia.org/wiki/The_Market_for_Lemons

Reputation systems Reputation systems• Significant number of reputation mechanisms Significant number of reputation mechanisms [Dellarocas et al, 2007]• Link analysis techniques [TrustRank, EigenTrust, NodeRanking, NetProbe, Snare] often applicable f li bl

Challenges in the Design of Reputation Systems • Insufficient participation p p • Overwhelmingly positive feedback • Dishonest reports • Identity changes Identity changes • Value imbalance exploitation (“milking the reputation”)

Insufficient Participation • Free‐riding: feedback constitutes a public good. Once available, everyone can costless‐ly benefit from it. • Disadvantage of early evaluators: provision of feedback p presupposes that the rater will assume the risks of transacting with pp g the ratee (competitive advantage to others). • [A [Avery et al. 1999] propose a mechanism whereby early l 1999] h i h b l evaluators are paid to provide information and later evaluators pay to balance the budget.

Overwhelmingly Positive Feedback (I) More than 99% of all feedbacks posted on eBay are positive. H I i d f 16% f ll However, Internet auctions accounted for 16% of all consumer fraud complaints received by the Federal Trade Commission in 2004. (http://www.consumer .gov/sentinel/) Reporting Bias The perils of reciprocity: • Reciprocity: Seller evaluated buyer, buyer evaluates seller • Exchange of courtesies • Positive reciprocity: positive ratings are given in the hope of getting a positive rating in return • Negative reciprocity: negative ratings are avoided because of fear of retaliation from the other party

Overwhelmingly Positive Feedback (II) “The sound of silence”: No news, bad news… • [Dellarocas and Wood 2008] Explore the frequency of different feedback patterns and use the non‐reports to compensate for reporting bias. • eBay traders are more likely to post feedback when satisfied than when dissatisfied than when dissatisfied • Support presence of positive and negative reciprocation among eBay traders.

Dishonest Reports • “Ballot stuffing” (unfairly high ratings): a seller colludes with a group of buyers in order to be given unfairly high ratings by them. group of buyers in order to be given unfairly high ratings by them • “Bad‐mouthing” (unfairly low ratings): Sellers can collude with buyers in order to “bad‐mouth” other sellers that they want to drive y y out the market. • Design incentive‐compatible mechanism to elicit honest feedbacks [ [Jurca and Faltings 2003: pay rater if report matches next; g p y p ; Miller et al. 2005: use a proper scoring rule to price value of report; Papaioannou and Stamoulis 2005: delay next transaction over time] • U “l t t l ” Use “latent class” models described earlier in the tutorial d l d ib d li i th t t i l (reputation systems is a form of crowdsourcing after all…)

Identity Changes • “Cheap pseudonyms”: easy to disappear and re‐ pp y y pp register under a new identity with almost zero cost. [Friedman and Resnick 2001] • I Introduce opportunities to misbehave without paying d ii ib h ih i reputational consequences. • Increase the difficulty of online identity changes • Impose upfront costs to new entrants: allow new identities (forget the past) but make it costly to create them

Value Imbalance Exploitation Three men attempt to sell a fake painting on eBay for $US 135,805. The sale was abandoned just prior to purchase when 135 805 Th l b d dj i h h the buyer became suspicious.(http://news.cnet.com/2100‐ 1017‐253848.html) • Reputation can be seen as an asset not only to Reputation can be seen as an asset, not only to promote oneself, but also as something that can be cashed in through a fraudulent transaction with high g g gain. “The Market for Evaluations”

The Market for Positive Feedbacks A selling strategy that eBay users are actually using the g gy y y g feedback market for gains in other markets. “Riddle for a PENNY! No shipping‐Positive Feedback” Riddle for a PENNY! No shipping Positive Feedback • 29‐cent loss even in the event of a successful sale • Price low, speed feedback accumulationPossible solutions: • Make the details of the transaction (besides the feedback itself) visible to other users • T Transaction‐weighted reputational statistics ti i ht d t ti l t ti ti [Brown 2006]

Challenges for Crowdsourcing Markets (I)• Two‐sided opportunistic behavior • Reciprocal systems worse than one‐side evaluation. In e‐commerce markets, only sellers are likely to behave opportunistically. No need for markets only sellers are likely to behave opportunistically No need for reciprocal evaluation! • In crowdsourcing markets, both sides can be fraudulent. Reciprocal systems are fraught with problems, though!• I Imperfect monitoring and heavy‐tailed participation f t it i dh t il d ti i ti • In e‐commerce markets, buyers can assess the product quality directly upon receiving. • In crowdsourcing markets, verifying the answers is sometimes as costly as providing them. • Sampling often does not work due to heavy tailed participation Sampling often does not work, due to heavy tailed participation distribution (lognormal, according to self‐reported surveys)

Challenges for Crowdsourcing Markets (II) • Constrained capacity of workers • In e‐commerce markets sellers usually have unlimited supply of In e commerce markets, sellers usually have unlimited supply of products. • In crowdsourcing, workers have constrained capacity (cannot be recommended continuously) • No “price premium” for high quality workers No “price premium” for high‐quality workers • In e‐commerce markets, sellers with high reputation can sell their products at a relatively high price (premium). • In crowdsourcing, it is the requester who set the prices, which are generally the same for all the workers.

The Importance of Task Discovery Heavy tailed distribution of completion times. Why?• Heavy tailed distribution of completion times Why? [Ipeirotis, “Analyzing the Amazon Mechanical Turk marketplace”, XRDS 2010]

The Importance and Danger of PrioritiesThe Importance and Danger of Priorities• [Barabasi, Nature 2005] showed that human actions [ , ] have power‐law completion times – Mainly result of prioritization – Wh t k When tasks ranked by priorities, power‐law results k db i iti l lt• [Cobham 1954] If queuing system completes tasks [Cobham, 1954] If queuing system completes tasks with two priority queues, and λ=μ, then power‐law completion times• [Chilton et al., HCOMP 2010] Workers on Turk pick tasks from most HITs or most recent queues tasks from “most HITs” or “most recent” queues

The UI hurts the market!• Practitioners know that HITs in 3rd page and after, p g , are not picked by workers. • Many such HITs are left to expire after months Many such HITs are left to expire after months, never completed.• Badly designed task discovery interface hurts every participant in the market! (and the reason for scientific modeling…) ti i t i th k t!• Better modeling as a queuing system may demonstrate other such improvements

Expert Search Expert Search• Find the best worker for a task (or within a task) Find the best worker for a task, (or within a task)• For a task: k – Significant amount of research in the topic of expert search [TREC track; Macdonald and Ounis, 2006] h – Check quality of workers across tasks http://url‐annotator.appspot.com/Admin/WorkersReport http://url annotator appspot com/Admin/WorkersReport• Within a task: [Donmez et al., 2009; Welinder, 2010] t a tas : [ o e et a , 009; e de , 0 0]

Directions for future research Directions for future research• Optimize allocation of tasks to worker based on completion time and expected quality• Explicitly take into consideration competition in market and Explicitly take into consideration competition in market and switch task for worker only when benefit outweighs switching overhead (task switching in CPU from O/S)• Recommender system for tasks (“workers like you performed well in…”)• Create a market with dynamic pricing for tasks, following the pricing model of the stock market (prices increase for task when work supply low, and vice versa) task when work supply low and vice versa)

Human Computation• Humans are not perfect mathematical models Humans are not perfect mathematical models• They exhibit noisy, stochastic behavior… h hibi i h i b h i• And exhibit common and systematic biases

Score the following from 1 to 10 1: not particularly bad or wrong 1 t ti l l b d 10: extremely evil a) Stealing a towel from a hotel b) Keeping a dime you ﬁnd on the ground p g y gc) Poisoning a barking dog [Parducci, 1968]

Score the following from 1 to 10 1: not particularly bad or wrong 1 t ti l l b d 10: extremely evil a) Testifying falsely for pay gg gb) Using guns on striking workersc) Poisoning a barking dog [Parducci, 1968]

Anchoring Anchoring• “Humans start with a first approximation (anchor) and f pp ( ) then make adjustments to that number based on additional information.” [Tversky & Kahneman, 1974]• [Paolacci et al, 2010] – Q1a: More or less than 65 African countries in UN? Q1a: More or less than 65 African countries in UN? – Q1b: More or less than 12 African countries in UN? – Q2: How many countries in Africa? – Group A mean: 42.6 – Group B mean: 18 5 Group B mean: 18.5

Anchoring Anchoring• Write down the last digit of their social security g f y number before placing bid for wine bottles. Users with lower SSN numbers bid lower…• In the Netﬂix contest, user with high ratings early on, biased towards higher ratings later in a on, biased towards higher ratings later in a session…• Crowdsourcing tasks can be affected by anchoring. [Moren et al, NIPS 2010] describe techniques for removing effects techniques for removing effects

Priming• Exposure to one stimulus influences another Exposure to one stimulus influences another• Stereotypes: – Asian americans perform better in math Asian‐americans perform better in math – Women perform worse in math• [Shih et al., 1999] asked Asian‐American women: – Q ti Questions about race: They did better in math test b t Th did b tt i th t t – Questions about gender: They did worse in math test

Exposure Effect Exposure Effect• Familiarity leads to liking Familiarity leads to liking...• [S [Stone and Alonso, 2010]: Evaluators of Bing d l 20 0] l f i search engine increase their ratings of relevance over time, for the same results l i f h l

Framing• Presenting the same option in different Presenting the same option in different formats leads to different formats. People avert options that imply loss [Tversky and avert options that imply loss [Tversky and Kahneman (1981)]

Framing: 600 people affected by deadly disease 600 people affected by deadly diseaseRoom 1a) save 200 peoples lives save 200 people s livesb) 33% chance of saving all 600 people and a 66% chance saving no one• 72% of participants chose option A 72% of participants chose option A• 28% of participants chose option BRoom 2Room 2c) 400 people died) 33% chance that no people will die; a 66% chance that all 600 will die• 78% of participants chose option D (equivalent to option B)• 22% of participants chose option C (equivalent to option A) People avert options that imply loss

Very long list of cognitive biases… Very long list of cognitive biases…• http://en.wikipedia.org/wiki/List_of_cognitive_biases p // p g/ / g• [Mozer et al., 2010] try to learn and remove sequential effects from human computation data…

Games with a Purpose [Luis von Ahn and Laura Dabbish, CACM 2008]Three generic game structures• Output agreement: – Type same output• I Input agreement: – Decide if having same input• Inversion problem: Inversion problem: – P1 generates output from input – P2 looks at P1‐output and guesses P1‐input

Output Agreement: ESP Game Output Agreement: ESP Game• Players look at common input Players look at common input• Need to agree on output

Improvements• Game‐theoretic analysis indicates that players Game theoretic analysis indicates that players will converge to easy words [Jain and Parkes]• Solution 1: Add “Taboo words” to prevent Solution 1: Add Taboo words to prevent guessing easy words• S l i 2 Ki Ki B Solution 2: KissKissBan, third player tries to hi d l i guess (and block) agreement

Input Agreement: TagATune p g g• Sometimes difficult to type identical output (e.g., “describe this song”) p ,• Show same of different input, let users describe, ask players if they have same input

Inversion Problem: Peekaboom Inversion Problem: Peekaboom• Non symmetric players Non‐symmetric players• Input: Image with word• Player 1 slowly reveals pic l l l l i• Player 2 tries to guess word

BUSHHINT

Protein folding Protein folding• Protein folding: Proteins fold from long chains into small balls, each in a very specific shape• Shape is the lower energy setting which the most Shape is the lower‐energy setting, which the most stable• Fold shape is very important to understand interactions with out molecules• Extremely expensive computationally! (too many degrees of freedom)

FoldIt Game• Humans are very good at reducing the search Humans are very good at reducing the search space• Humans try to fold the protein into a minimal energy state. • Can leave protein unfinished and let others try from there…

Case Study: FreebaseCase Study: FreebasePraveen Paritosh, Google

Crowdsourcing Case Study AdSafe

A few of the tasks in the past• Detect pages that discuss swine flu – Pharmaceutical firm had drug “treating” (off-label) swine flu – FDA prohibited pharmaceutical company to display drug ad in pages about swine flu – Two days to build and go live• Big fast-food chain does not want ad to appear: – In pages that discuss the brand (99% negative sentiment) – In pages discussing obesity – Three days to build and go live 180

Need to build models fast Need to build models fast • T diti Traditionally, modeling teams have invested substantial ll d li t h i t d b t ti l internal resources in data formulation, information extraction, cleaning, and other preprocessing No time for such things… • However, now, we can outsource preprocessing tasks, such as labeling, feature extraction, verifying information l b li f t t ti if i i f ti extraction, etc. – using Mechanical Turk, oDesk, etc. – quality may be lower than expert labeling (much?) – but low costs can allow massive scale181

AdSafe workflow • Find URLs for a given topic (hate speech, gambling, alcohol abuse, guns, bombs, celebrity gossip, etc etc) b b b l bi i ) http://url‐collector.appspot.com/allTopics.jsp • Classify URLs into appropriate categories Classify URLs into appropriate categories http://url‐annotator.appspot.com/AdminFiles/Categories.jsp • Mesure quality of the labelers and remove spammers http://qmturk.appspot.com/ htt // t k t / • Get humans to “beat” the classifier by providing cases where the classifier fails http://adsafe‐beatthemachine.appspot.com/182

Case Study: OCR and ReCAPTCHACase Study: OCR and ReCAPTCHA

Scaling Crowdsourcing: Use Machine Learning Need to scale crowdsourcing Basic idea: Build a machine learning model and use it instead of humansNew case Automatic Model Automatic (through machine learning) Answer Existing data (through crowdsourcing)

Scaling Crowdsourcing: Iterative training Ti Triage: – machine when confident – humans when not confident Retrain using the new human input Automatic → improve model Answer → reduce need for human inputNew Case Automatic Model (through machine learning) Data from existing Get human(s) to answer crowdsourced answers

Scaling Crowdsourcing: Iterative training, with noise Machine when confident, humans otherwise Ask as many humans as necessary to ensure quality Automatic AnswerNew Case Automatic Model Not confident (through machine l (th h hi learning) i ) for quality? Data from existing Get human(s) to crowdsourced answers answer Confident for quality?

Scaling Crowdsourcing: Iterative training, with noise Machine when confident, humans otherwise Ask as many humans as necessary to ensure quality – Or even get other machines… machines Automatic AnswerNew Case Automatic Model Not confident (through machine l (th h hi learning) i ) about quality? Data from existing Get human(s) or crowdsourced answers other machines Confident about quality? to answer

Example: ReCAPTCHA + Google Books portion distinguished Fixes errors of Optical Character Recognition (OCR ~ 1% error rate 20% rate, 20%- 30% for 18th and 19th century books, according to today’s NY Times article) Improves further the OCR algorithm, reducing error rate “40 million R CAPTCHA every d ” (2008) Fi i 40 000 b k a day illi ReCAPTCHAs day” Fixing 40,000 books d – [Unofficial quote from Luis]: 400M/day (2010) – All books ever written:100 million books (~12yrs??)

Thank you!Thank you!

References• David Alan Grier: When Computers Were Human, Princeton University Press, 2005• [ESP Game] Luis von Ahn and Laura Dabbish, Labeling images with a computer game, CHI 2004 [ESP Game] Luis von Ahn and Laura Dabbish Labeling images with a computer game CHI 2004• [ImageNet] http://www.image‐net.org/about‐publication• O. Alonso, G. Kazai, and S. Mizzaro. Crowdsourcing for search engine evaluation, Springer 2011• Kuncheva et al., Limits on the majority vote accuracy in classiﬁer fusion, Pattern Analysis and Applications, 2003• Robert T. Clemen, Robert L. Winkler, Limits for the Precision and Value of Information from Dependent Sources, Robert T Clemen Robert L Winkler Limits for the Precision and Value of Information from Dependent Sources Operations Research, 1985• Robert T. Clemen and Robert L. Winkler, Unanimity and Compromise among Probability Forecasters, Management Science, 1990• J. Whitehill, P. Ruvolo, T. Wu, J. Bergsma, and J. Movellan. Whose vote should count more: Optimal integration J. Whitehill, P. Ruvolo, T. Wu, J. Bergsma, and J. Movellan. Whose vote should count more: Optimal integration of labels from labelers of unknown expertise. In NIPS, 2009• Peter Welinder, Steve Branson, Serge Belongie, Pietro Perona, The Multidimensional Wisdom of Crowds, In NIPS, 2010• Wim J. van der Linden, Ronald K. Hambleton, Handbook of modern item response theory, Springer 1997• Uebersax, A latent trait finite mixture model for the analysis of rating agreement, Biometrics 1993• Uebersax, Statistical Modeling of Expert Ratings on Medical Treatment Appropriateness, JASA 1993• Wang, Ipeirotis, Provost, Managing Crowdsourced Workers, Workshop on Business Intelligence 2011• , y g y g p g, Donmez et al, Efﬁciently Learning the Accuracy of Labeling Sources for Selective Sampling, KDD 2009• Donmez et al., A Probabilistic Framework to Learn from Multiple Annotators with Time‐Varying Accuracy, SDM 2010

References• Ariely et al., The Effects of Averaging Subjective Probability Estimates Between and Within Judges, Journal of Experimental Psychology 2000 Journal of Experimental Psychology 2000• Clemen and Winkler Combining probability distributions from experts in risk analysis, Risk Analysis 1999• Burr Settles, Active Learning Literature Survey, Computer Science Technical Report 1648, University of Wisconsin‐Madison• Kumar and Lease, Modeling Annotator Accuracies for Supervised Learning, CSDM 2011 Kumar and Lease Modeling Annotator Accuracies for Supervised Learning CSDM 2011• Ipeirotis et al., Repeated Labeling Using Multiple Noisy Labelers, NYU Working Paper CeDER‐10‐03• Carla E. Brodley , Mark A. Friedl, Identifying Mislabeled Training Data JAIR 1999• Little, Chilton, Goldman, Miller: TurKit: human computation algorithms on mechanical turk. UIST 2010• Dai, Mausam, Weld, Decision‐Theoretic Control of Crowd‐Sourced Workﬂows, AAAI 2010• Kern, Thies, and Satzger, Statistical Quality Control for Human‐based Electronic Services, ICSOC 2010• Montgomery, Introduction to statistical quality control, Wiley 2007• Aniket Kittur, Boris Smus, Robert E. Kraut, CrowdForge: Crowdsourcing Complex Work, CHI 2011 , , , g g p ,• Anand Kulkarni, Matthew Can, Turkomatic: Automatic Recursive Task Design for Mechanical Turk, CHI2011• Haoqi Zhang, Eric Horvitz, Rob C. Miller, David C. Parkes, Crowdsourcing General Computation, CHI g crowdsourcing workshop 2011p• Eric Huang, Haoqi Zhang, David Parkes, Krzysztof Gajos, and Yiling Chen, Toward Automatic Task Design: A Progress Report, HCOMP 2010

References• Mason and Watts, Financial Incentives and the “Performance of Crowds”, HCOMP 2009• Yan, Kumar, Ganesan, CrowdSearch: Exploiting Crowds for Accurate Real‐time Image Search on Mobile Yan Kumar Ganesan “CrowdSearch: Exploiting Crowds for Accurate Real time Image Search on Mobile Phones”, MobiSys 2010• Ipeirotis, Analyzing the Mechanical Turk Marketplace, XRDS 2010• Wang, Faridani, Ipeirotis, Estimating Completion Time for Crowdsourced Tasks Using Survival Analysis Models. CSDM 2010 Models CSDM 2010• Chilton et al, Task search in a human computation market, HCOMP 2010• Bingham et al, VizWiz: nearly real‐time answers to visual questions, UIST 2011• Horton and Chilton: The labor economics of paid crowdsourcing. EC 2010• Ariely et al., Large stakes and big mistakes, Review of Economic Studies, 2009• Huang et al., Toward Automatic Task Design: A Progress Report, HCOMP 2010• Quinn, Bederson, Yeh, Lin.: CrowdFlow: Integrating Machine Learning with Mechanical Turk for Speed‐ Cost‐Quality Flexibility• Franklin et al.: Answering Queries with Crowdsourcing, SIGMOD 2011• Parameswaran et al.: Human‐assisted Graph Search: Its Okay to Ask Questions, VLDB 2011• Mitzenmacher, An introduction to human‐guided search, XRDS 2010• Marcus et al, Crowdsourced Databases: Query Processing with People, CIDR 2011 Marcus et al Crowdsourced Databases: Query Processing with People CIDR 2011• Deng, Berg, Li and Fei‐Fei. What does classifying more than 10,000 image categories tell us? ECCV 2010• Kleinberg, Navigation in a small world, Nature 2000

References• Lugosi, Learning with an unreliable teacher. Pattern Recognition 1992• Smyth, Learning with probabilistic supervision. Selecting Good Models, MIT Press, Apr. 1995 h i ih b bili i ii l i d d l• Smyth, Bounds on the mean classification error rate of multiple experts, Pattern Recognition Let. 1996• Gneezy and Rustichini, Pay enough or dont pay at all, The Quarterly Journal of Economics, 2000• Camerer, et al. Labor supply of New York City cabdrivers: One day at a time. The Quarterly Journal of Camerer et al Labor supply of New York City cabdrivers: One day at a time The Quarterly Journal of Economics, 1997• Farber. Reference‐dependent preferences and labor supply: The case of New York City taxi drivers. American Economic Review, 2008.• Fehr and Goette. Do workers work more if wages are high?: Evidence from a randomized field F h dG D k k if hi h? E id f d i d fi ld experiment. American Economic Review, 2007• Chandler and Kepelner, Breaking Monotony with Meaning: Motivation in Crowdsourcing Markets , 2010• Ariely, Kamenica, Prelec. Man’s Search for Meaning: The Case of Legos. Journal of Economic Behavior & y, , g g Organization 2008• Farmer and Glass, Building Web Reputation Systems, O’Reilly 2010• Raykar, Yu, Zhao, Valadez, Florin, Bogoni, and Moy. Learning from crowds. JMLR 2010.• Mason and Watts, Collective problem solving in networks, 2011 M d W tt C ll ti bl l i i t k 2011• Dellarocas, Dini and Spagnolo. Designing Reputation Mechanisms. Chapter 18 in Handbook of Procurement, Cambridge University Press, 2007

References• Mozer et al, Decontaminating Human Judgments by Removing Sequential Dependencies, NIPS 2010• Tversky, & Kahneman, Judgment under uncertainty: Heuristics and biases. Science, 1974• Parducci. The relativism of absolute judgment. Scientiﬁc American, 1968 d i h l i i f b l j d i iﬁ i• Paolacci, Chandler, Ipeirotis. Running Experiments using Amazon Mechanical Turk. Judgment and Decision Making, 2010• Shih, Pittinsky, Ambady, Stereotype susceptibility: Identity salience and shifts in quantitative , y, y, yp p y y q performance, Psychological Science, 1999• Bernstein et al, Soylent: A Word Processor with a Crowd Inside, UIST 2010• Malone, Heuristics for designing enjoyable user interfaces: Lessons from computer games. , 1982,• Malone, What makes things fun to learn? Heuristics for designing instructional computer games. 1980 M l Wh k hi f l ?H i i f d i i i i l 1980• Luis von Ahn, Laura Debbish, Designing Games with a Purpose, CACM 2008• Jain and Parkes, A Game‐Theoretic Analysis of the ESP Game, Working Paper • Cooper et al, Predicting protein structures with a multiplayer online game, Nature, August 2010 Cooper et al Predicting protein structures with a multiplayer online game Nature August 2010

Books, Courses, and Surveys Thanks to Matt Lease http://ir.ischool.utexas.edu/crowd/ Thanks to Matt Lease http://ir.ischool.utexas.edu/crowd/• Omar Alonso, Gabriella Kazai, and Stefano Mizzaro. Crowdsourcing for Search Engine Evaluation: Why and How. To be published by Springer, 2011.• Edith Law and Luis von Ahn. Human Computation: An Integrated Approach to Learning Edith Law and Luis von Ahn Human Computation: An Integrated Approach to Learning from the Crowd. To be published in Morgan & Claypool Synthesis Lectures on Artificial Intelligence and Machine Learning, 2011. http://www.morganclaypool.com/toc/aim/1/1• Deepak Ganesan. CS691CS: Crowdsourcing ‐ Opportunities & Challenges (Fall 2010). UMass Amherst http://www.cs.umass.edu/~dganesan/courses/fall10• Matt Lease. CS395T/INF385T: Crowdsourcing: Theory, Methods, and Applications (Spring 2011). UT Austin. http://courses.ischool.utexas.edu/Lease_Matt/2011/Spring/INF385T• Yuen, Chen, King: A Survey of Human Computation Systems, SCA 2009• Quinn, Bederson: A Taxonomy of Distributed Human Computation, CHI 2011• Doan, Ramakrishnan, Halevy: Crowdsourcing Systems on the World‐Wide Web, CACM 2011

Tutorials Thanks to Matt Lease http://ir.ischool.utexas.edu/crowd/ Thanks to Matt Lease http://ir.ischool.utexas.edu/crowd/• AAAI 2011 (w HCOMP 2011): Human Computation: Core Research Questions and State of the Art (Edith Law and Luis von Ahn)• WSDM 2011: Crowdsourcing 101: Putting the WSDM of Crowds to Work for You (Omar WSDM 2011: Crowdsourcing 101: Putting the WSDM of Crowds to Work for You (Omar Alonso and Matthew Lease) http://ir.ischool.utexas.edu/wsdm2011_tutorial.pdf• LREC 2010 Tutorial: Statistical Models of the Annotation Process (Bob Carpenter and Massimo Poesio) http://lingpipe blog.com/2010/05/17/lrec 2010 tutorial modeling Massimo Poesio) http://lingpipe‐blog.com/2010/05/17/lrec‐2010‐tutorial‐modeling‐ data‐annotation/• ECIR 2010: Crowdsourcing for Relevance Evaluation. (Omar Alonso) http://wwwcsif.cs.ucdavis.edu/~alonsoom/crowdsourcing.html• CVPR 2010: Mechanical Turk for Computer Vision. (Alex Sorokin and Fei‐Fei Li) http://sites.google.com/site/turkforvision/• CIKM 2008: Crowdsourcing for Relevance Evaluation (Daniel Rose) http://videolectures.net/cikm08_rose_cfre/ htt // id l t t/ ik 08 f /

http://behind-the-enemy-lines.blogspot.com	532
http://www.behind-the-enemy-lines.com	470
http://feeds.feedburner.com	256
http://paper.li	47
http://www.slideshare.net	7
url_unknown	5
http://twitter.com	3
http://topsitemap.blogspot.com	3
http://www.ourashima.somee.com	2
http://tatsukii.tumblr.com	2
http://www.twylah.com	1
http://safe.tumblr.com	1
http://misoji.tumblr.com	1
https://twitter.com	1
http://us-w1.rockmelt.com	1
http://translate.googleusercontent.com	1

Managing Crowdsourced Human Computation: A Tutorial

by Panos Ipeirotis on Apr 05, 2011

Accessibility

Categories

Tags

Upload Details

Usage Rights

16 Embeds 1,333

Statistics

Managing Crowdsourced Human Computation: A Tutorial — Presentation Transcript