SlideShare una empresa de Scribd logo
1 de 19
21st International Conference on Cooperative Information Systems (CoopIS’13)
September 11-13, 2013, Graz, Austria
Ognjen Scekic, Christoph Dorn, Schahram Dustdar
Distributed Systems Group
Vienna University of Technology
http://dsg.tuwien.ac.at
Simulation-Based Modeling and Evaluation
of Incentive Schemes in Crowdsourcing
2 CoopIS’13
Outline
 Incentives in Crowdsourcing today and tomorrow
 Problems with evaluating incentives
 Our approach
– Simulation model
– Simulation methodology & tools
 Real-world scenario example
 Conclusion and Outlook
3 CoopIS’13
Incentives & Rewards
• Incentives
Stimulate (motivate) or discourage
certain worker activities before the
actual execution of those activities.
• Rewards
Any kind of recompense for worthy
services rendered or retribution for
wrongdoing exerted upon workers
after the completion of activity.
• Incentive Mechanism
A plan (rule) for assigning rewards.
4 CoopIS’13
Evolution of Crowdsourcing
Conventional workflows
• formal description
• structured execution
• predefined roles and activities
• complex tasks
Crowdsourcing
• simple tasks
• anonymous replaceable actors
• short, unstructured interactions
• No interaction/collaboration
among actors
+
=
Socio-technical Collective Adaptive Systems
• ad-hoc assembled teams
• complex tasks
• social orchestration
• indirect adaptation
5 CoopIS’13
 Incentive schemes can be built
by composing and customizing
well-known incentive elements.
 Programmable incentive management
 Portable, reusable, scalable incentives.
 Problem:
Composition  evaluation complexity
– How to prevent malicious workers?
– How to anticipate free riding, multitasking,
tragedy of the commons?
– How to assess appropriate reward amounts
Modeling Incentives – Problems
6 CoopIS’13
 Need a systematic approach in designing and evaluating
incentive schemes before deployment on real systems.
 How to select, customize and evaluate appropriate atomic
incentive mechanisms and how to compose them for a given
crowdsourcing scenario?
 We present:
– Simulation model of incentive mechanism
– Modeling and simulation methodology for approximate
estimation of the composition of incentive mechanisms.
Contributions
7 CoopIS’13
1) Mathematical incentive models
(e.g., principal-agent theory, game theory)
2) Empirical evaluation
Existing Evaluation Approaches
PRO CON
precise and reliable related to particular collaboration
patterns, cannot handle unforeseen
runtime changes
PRO CON
good for evaluating simple existing
incentives and behavioral responses
impossibility to isolate particular
mechanisms and their effects, or
causes of behavior in complex cases;
platform limitations (e.g.,
communication channels, predefined
incentives and metrics);
8 CoopIS’13
3) Experimental evaluation
(e.g., on micro-task platforms, with students, volunteers)
Existing Evaluation Approaches
PRO CON
controlled environment and
reproducible setups
platform limitations (e.g.,
communication channels, predefined
incentives and metrics);
limited monetary funds may derive
skewed results;
working with people inherently willing
or forced to perform work may derive
skewed results
9 CoopIS’13
 Offer methodology for quickly selecting, composing and
customizing existing incentive mechanisms.
 Roughly predicting effects of composition in dynamic
crowdsourcing environments.
 Model and simulation parameters can be changed dynamically,
allowing quick testing of different incentive scheme setups and
behavioral responses at low cost.
 Modeling of incentives and responses of arbitrary complexity.
 We do not devise novel nor optimal incentive mechanisms!
Our approach
10 CoopIS’13
Simulation Model of Incentive Mechanism
Decision-making function fa considers:
1) the statistically or intentionally determined
personality of the worker St
2) historical records of past actions {S0, … , St-1}
3) authority’s view of worker’s performance Mj
4) performance of other workers {Mk}, k ≠ j
5) promised rewards R
Incentive mechanism IM considers:
1) current state of artifact Ki
2) the current performance metrics of
worker Mj
3) output from another incentive mechanism
returning the same type of reward R′ak
11 CoopIS’13
 True power of incentives  composition of incentive mechanisms
 Two basic operators on incentive mechanisms:
– addition (+) and functional composition ()
– operate on common metrics
– final metrics’ values advertised to workers represent the promised reward
 Major difficulty in designing successful incentive strategies is to
properly choose performance metrics, basic incentive
mechanisms and the proper composition.
Simulating Complex Incentive Strategies
12 CoopIS’13
Simulation Methodology
1) Defining domain-specific meta-model by extending core meta-model
2) Capturing worker behavioral patterns and reward calculation into executable model
3) Defining scenarios, assumptions, and configurations for individual simulation runs
4) Evaluating and interpreting simulation results
1 2 3 4
13 CoopIS’13
Simulation Methodology
1) Defining domain-specific meta-model by extending core meta-model
2) Capturing worker behavioral patterns and reward calculation into executable model
3) Defining scenarios, assumptions, and configurations for individual simulation runs
4) Evaluating and interpreting simulation results
1 2
DomainPro
Designer
*www.quandarypeak.com
14 CoopIS’13
Simulation Methodology
1) Defining domain-specific meta-model by extending core meta-model
2) Capturing worker behavioral patterns and reward calculation into executable model
3) Defining scenarios, assumptions, and configurations for individual simulation runs
4) Evaluating and interpreting simulation results
3
DomainPro Analyst
4
*www.quandarypeak.com
15 CoopIS’13
 Simulation model of a realistic scenario, inspired by
– Citizen-driven traffic reporting (SmartJourney – Aberdeen)
– Crowdsourced software testing
 Generalized scenario:
– Entities: Authority, Workers, Situations, Reports
– Activities: Submit, Improve, Rate, Report duplicates
– Metrics: Reputation (for trustworthiness), Points (for productivity)
– Incentives: Three incentive mechanisms:
 IM1 – fixed amounts of points per activity
 IM2 – points related with report quality
 IM3 – users are assigned reputation based on past activities
Evaluation
16 CoopIS’13
 Composite Incentive Schemes (CIS) evaluated:
 3 Experiments:
– Exp1: Compare impact of CIS1, CIS2, and CIS3 on authority cost.
– Exp2: Analyze effects of having too few or too many
workers per situation
– Exp3: Evaluate effects of malicious workers (0-50%) on cost.
Evaluation
17 CoopIS’13
 CIS3 most reasonable to use. Can cope well with up to
20% of malicious workers.
Evaluation Results – Example
18 CoopIS’13
 Presented a methodology for modeling and simulating
incentives in crowdsourcing environments.
 Advantages:
– Useful for quick, runtime, approximate evaluations
of different compositions of incentive mechanisms.
 Drawbacks:
– Inconclusive results. See: Advantages
 Future Work:
– Devise suitable models for more complex socio-technical systems.
Conclusion & Outlook
21st International Conference on Cooperative Information Systems (CoopIS’13)
September 11-13, 2013, Graz, Austria
Ognjen Scekic, Christoph Dorn, Schahram Dustdar
Distributed Systems Group
Vienna University of Technology
http://dsg.tuwien.ac.at
Thank you!
Questions?
21st International Conference on Cooperative Information Systems (CoopIS’13)
September 11-13, 2013, Graz, Austria

Más contenido relacionado

Similar a Simulation-Based Modeling and Evaluation of Incentive Schemes in Crowdsourcing Environments

credit tk72 C o M M u n i C AT i o .docx
credit tk72    C o M M u n i C AT i o .docxcredit tk72    C o M M u n i C AT i o .docx
credit tk72 C o M M u n i C AT i o .docxfaithxdunce63732
 
IRJET - Job Portal Analysis and Salary Prediction System
IRJET -  	  Job Portal Analysis and Salary Prediction SystemIRJET -  	  Job Portal Analysis and Salary Prediction System
IRJET - Job Portal Analysis and Salary Prediction SystemIRJET Journal
 
Applicants Qualification Filtering System
Applicants Qualification Filtering SystemApplicants Qualification Filtering System
Applicants Qualification Filtering SystemSiti Nabilah Ismail
 
Ron's muri presentation
Ron's muri presentationRon's muri presentation
Ron's muri presentationgowinraj
 
Software requirement analysis enhancements byprioritizing re
Software requirement analysis enhancements byprioritizing reSoftware requirement analysis enhancements byprioritizing re
Software requirement analysis enhancements byprioritizing reAlleneMcclendon878
 
Evaluate deep q learning for sequential targeted marketing with 10-fold cross...
Evaluate deep q learning for sequential targeted marketing with 10-fold cross...Evaluate deep q learning for sequential targeted marketing with 10-fold cross...
Evaluate deep q learning for sequential targeted marketing with 10-fold cross...Jian Wu
 
Data Science for Business Managers - An intro to ROI for predictive analytics
Data Science for Business Managers - An intro to ROI for predictive analyticsData Science for Business Managers - An intro to ROI for predictive analytics
Data Science for Business Managers - An intro to ROI for predictive analyticsAkin Osman Kazakci
 
Tech meetup Data Driven - Codemotion
Tech meetup Data Driven - Codemotion Tech meetup Data Driven - Codemotion
Tech meetup Data Driven - Codemotion antimo musone
 
Unit 6 Simulation.pptx
Unit 6 Simulation.pptxUnit 6 Simulation.pptx
Unit 6 Simulation.pptxHafiz20006
 
Introduction to Machine Learning.pptx
Introduction to Machine Learning.pptxIntroduction to Machine Learning.pptx
Introduction to Machine Learning.pptxDr. Amanpreet Kaur
 
Introduction to System, Simulation and Model
Introduction to System, Simulation and ModelIntroduction to System, Simulation and Model
Introduction to System, Simulation and ModelMd. Hasan Imam Bijoy
 
Methodologies for the Development of Crowd and Social-based applications
Methodologies for the Development of Crowd and Social-based applicationsMethodologies for the Development of Crowd and Social-based applications
Methodologies for the Development of Crowd and Social-based applicationsAndrea Mauri
 
IRJET- Evaluation Technique of Student Performance in various Courses
IRJET- Evaluation Technique of Student Performance in various CoursesIRJET- Evaluation Technique of Student Performance in various Courses
IRJET- Evaluation Technique of Student Performance in various CoursesIRJET Journal
 
Continuous Evaluation of Collaborative Recommender Systems in Data Stream Man...
Continuous Evaluation of Collaborative Recommender Systems in Data Stream Man...Continuous Evaluation of Collaborative Recommender Systems in Data Stream Man...
Continuous Evaluation of Collaborative Recommender Systems in Data Stream Man...Dr. Cornelius Ludmann
 
Predictive analytics in financial service
Predictive analytics in financial servicePredictive analytics in financial service
Predictive analytics in financial servicePrasad Narasimhan
 
Analysis on Fraud Detection Mechanisms Using Machine Learning Techniques
Analysis on Fraud Detection Mechanisms Using Machine Learning TechniquesAnalysis on Fraud Detection Mechanisms Using Machine Learning Techniques
Analysis on Fraud Detection Mechanisms Using Machine Learning TechniquesIRJET Journal
 
Crowdsourcing predictors of behavioral outcomes
Crowdsourcing predictors of behavioral outcomesCrowdsourcing predictors of behavioral outcomes
Crowdsourcing predictors of behavioral outcomesJPINFOTECH JAYAPRAKASH
 
Autonomy Incubator Seminar Series: Tractable Robust Planning and Model Learni...
Autonomy Incubator Seminar Series: Tractable Robust Planning and Model Learni...Autonomy Incubator Seminar Series: Tractable Robust Planning and Model Learni...
Autonomy Incubator Seminar Series: Tractable Robust Planning and Model Learni...AutonomyIncubator
 
Automated Feature Selection and Churn Prediction using Deep Learning Models
Automated Feature Selection and Churn Prediction using Deep Learning ModelsAutomated Feature Selection and Churn Prediction using Deep Learning Models
Automated Feature Selection and Churn Prediction using Deep Learning ModelsIRJET Journal
 

Similar a Simulation-Based Modeling and Evaluation of Incentive Schemes in Crowdsourcing Environments (20)

credit tk72 C o M M u n i C AT i o .docx
credit tk72    C o M M u n i C AT i o .docxcredit tk72    C o M M u n i C AT i o .docx
credit tk72 C o M M u n i C AT i o .docx
 
IRJET - Job Portal Analysis and Salary Prediction System
IRJET -  	  Job Portal Analysis and Salary Prediction SystemIRJET -  	  Job Portal Analysis and Salary Prediction System
IRJET - Job Portal Analysis and Salary Prediction System
 
Applicants Qualification Filtering System
Applicants Qualification Filtering SystemApplicants Qualification Filtering System
Applicants Qualification Filtering System
 
Ron's muri presentation
Ron's muri presentationRon's muri presentation
Ron's muri presentation
 
Software requirement analysis enhancements byprioritizing re
Software requirement analysis enhancements byprioritizing reSoftware requirement analysis enhancements byprioritizing re
Software requirement analysis enhancements byprioritizing re
 
Evaluate deep q learning for sequential targeted marketing with 10-fold cross...
Evaluate deep q learning for sequential targeted marketing with 10-fold cross...Evaluate deep q learning for sequential targeted marketing with 10-fold cross...
Evaluate deep q learning for sequential targeted marketing with 10-fold cross...
 
Data Science for Business Managers - An intro to ROI for predictive analytics
Data Science for Business Managers - An intro to ROI for predictive analyticsData Science for Business Managers - An intro to ROI for predictive analytics
Data Science for Business Managers - An intro to ROI for predictive analytics
 
Tech meetup Data Driven - Codemotion
Tech meetup Data Driven - Codemotion Tech meetup Data Driven - Codemotion
Tech meetup Data Driven - Codemotion
 
Unit 6 Simulation.pptx
Unit 6 Simulation.pptxUnit 6 Simulation.pptx
Unit 6 Simulation.pptx
 
Introduction to Machine Learning.pptx
Introduction to Machine Learning.pptxIntroduction to Machine Learning.pptx
Introduction to Machine Learning.pptx
 
Introduction to System, Simulation and Model
Introduction to System, Simulation and ModelIntroduction to System, Simulation and Model
Introduction to System, Simulation and Model
 
Methodologies for the Development of Crowd and Social-based applications
Methodologies for the Development of Crowd and Social-based applicationsMethodologies for the Development of Crowd and Social-based applications
Methodologies for the Development of Crowd and Social-based applications
 
IRJET- Evaluation Technique of Student Performance in various Courses
IRJET- Evaluation Technique of Student Performance in various CoursesIRJET- Evaluation Technique of Student Performance in various Courses
IRJET- Evaluation Technique of Student Performance in various Courses
 
SIMULATION
SIMULATIONSIMULATION
SIMULATION
 
Continuous Evaluation of Collaborative Recommender Systems in Data Stream Man...
Continuous Evaluation of Collaborative Recommender Systems in Data Stream Man...Continuous Evaluation of Collaborative Recommender Systems in Data Stream Man...
Continuous Evaluation of Collaborative Recommender Systems in Data Stream Man...
 
Predictive analytics in financial service
Predictive analytics in financial servicePredictive analytics in financial service
Predictive analytics in financial service
 
Analysis on Fraud Detection Mechanisms Using Machine Learning Techniques
Analysis on Fraud Detection Mechanisms Using Machine Learning TechniquesAnalysis on Fraud Detection Mechanisms Using Machine Learning Techniques
Analysis on Fraud Detection Mechanisms Using Machine Learning Techniques
 
Crowdsourcing predictors of behavioral outcomes
Crowdsourcing predictors of behavioral outcomesCrowdsourcing predictors of behavioral outcomes
Crowdsourcing predictors of behavioral outcomes
 
Autonomy Incubator Seminar Series: Tractable Robust Planning and Model Learni...
Autonomy Incubator Seminar Series: Tractable Robust Planning and Model Learni...Autonomy Incubator Seminar Series: Tractable Robust Planning and Model Learni...
Autonomy Incubator Seminar Series: Tractable Robust Planning and Model Learni...
 
Automated Feature Selection and Churn Prediction using Deep Learning Models
Automated Feature Selection and Churn Prediction using Deep Learning ModelsAutomated Feature Selection and Churn Prediction using Deep Learning Models
Automated Feature Selection and Churn Prediction using Deep Learning Models
 

Último

Factors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptxFactors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptxKatpro Technologies
 
Powerful Google developer tools for immediate impact! (2023-24 C)
Powerful Google developer tools for immediate impact! (2023-24 C)Powerful Google developer tools for immediate impact! (2023-24 C)
Powerful Google developer tools for immediate impact! (2023-24 C)wesley chun
 
What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?Antenna Manufacturer Coco
 
Exploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone ProcessorsExploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone Processorsdebabhi2
 
GenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day PresentationGenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day PresentationMichael W. Hawkins
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsMaria Levchenko
 
Breaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path MountBreaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path MountPuma Security, LLC
 
🐬 The future of MySQL is Postgres 🐘
🐬  The future of MySQL is Postgres   🐘🐬  The future of MySQL is Postgres   🐘
🐬 The future of MySQL is Postgres 🐘RTylerCroy
 
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024The Digital Insurer
 
Presentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreterPresentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreternaman860154
 
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc
 
08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking MenDelhi Call girls
 
Boost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfBoost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfsudhanshuwaghmare1
 
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdfThe Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdfEnterprise Knowledge
 
Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024The Digital Insurer
 
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptx
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptxEIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptx
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptxEarley Information Science
 
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Miguel Araújo
 
Advantages of Hiring UIUX Design Service Providers for Your Business
Advantages of Hiring UIUX Design Service Providers for Your BusinessAdvantages of Hiring UIUX Design Service Providers for Your Business
Advantages of Hiring UIUX Design Service Providers for Your BusinessPixlogix Infotech
 
From Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time AutomationFrom Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time AutomationSafe Software
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...apidays
 

Último (20)

Factors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptxFactors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptx
 
Powerful Google developer tools for immediate impact! (2023-24 C)
Powerful Google developer tools for immediate impact! (2023-24 C)Powerful Google developer tools for immediate impact! (2023-24 C)
Powerful Google developer tools for immediate impact! (2023-24 C)
 
What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?
 
Exploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone ProcessorsExploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone Processors
 
GenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day PresentationGenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day Presentation
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed texts
 
Breaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path MountBreaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path Mount
 
🐬 The future of MySQL is Postgres 🐘
🐬  The future of MySQL is Postgres   🐘🐬  The future of MySQL is Postgres   🐘
🐬 The future of MySQL is Postgres 🐘
 
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
 
Presentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreterPresentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreter
 
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
 
08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men
 
Boost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfBoost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdf
 
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdfThe Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
 
Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024
 
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptx
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptxEIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptx
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptx
 
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
 
Advantages of Hiring UIUX Design Service Providers for Your Business
Advantages of Hiring UIUX Design Service Providers for Your BusinessAdvantages of Hiring UIUX Design Service Providers for Your Business
Advantages of Hiring UIUX Design Service Providers for Your Business
 
From Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time AutomationFrom Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time Automation
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
 

Simulation-Based Modeling and Evaluation of Incentive Schemes in Crowdsourcing Environments

  • 1. 21st International Conference on Cooperative Information Systems (CoopIS’13) September 11-13, 2013, Graz, Austria Ognjen Scekic, Christoph Dorn, Schahram Dustdar Distributed Systems Group Vienna University of Technology http://dsg.tuwien.ac.at Simulation-Based Modeling and Evaluation of Incentive Schemes in Crowdsourcing
  • 2. 2 CoopIS’13 Outline  Incentives in Crowdsourcing today and tomorrow  Problems with evaluating incentives  Our approach – Simulation model – Simulation methodology & tools  Real-world scenario example  Conclusion and Outlook
  • 3. 3 CoopIS’13 Incentives & Rewards • Incentives Stimulate (motivate) or discourage certain worker activities before the actual execution of those activities. • Rewards Any kind of recompense for worthy services rendered or retribution for wrongdoing exerted upon workers after the completion of activity. • Incentive Mechanism A plan (rule) for assigning rewards.
  • 4. 4 CoopIS’13 Evolution of Crowdsourcing Conventional workflows • formal description • structured execution • predefined roles and activities • complex tasks Crowdsourcing • simple tasks • anonymous replaceable actors • short, unstructured interactions • No interaction/collaboration among actors + = Socio-technical Collective Adaptive Systems • ad-hoc assembled teams • complex tasks • social orchestration • indirect adaptation
  • 5. 5 CoopIS’13  Incentive schemes can be built by composing and customizing well-known incentive elements.  Programmable incentive management  Portable, reusable, scalable incentives.  Problem: Composition  evaluation complexity – How to prevent malicious workers? – How to anticipate free riding, multitasking, tragedy of the commons? – How to assess appropriate reward amounts Modeling Incentives – Problems
  • 6. 6 CoopIS’13  Need a systematic approach in designing and evaluating incentive schemes before deployment on real systems.  How to select, customize and evaluate appropriate atomic incentive mechanisms and how to compose them for a given crowdsourcing scenario?  We present: – Simulation model of incentive mechanism – Modeling and simulation methodology for approximate estimation of the composition of incentive mechanisms. Contributions
  • 7. 7 CoopIS’13 1) Mathematical incentive models (e.g., principal-agent theory, game theory) 2) Empirical evaluation Existing Evaluation Approaches PRO CON precise and reliable related to particular collaboration patterns, cannot handle unforeseen runtime changes PRO CON good for evaluating simple existing incentives and behavioral responses impossibility to isolate particular mechanisms and their effects, or causes of behavior in complex cases; platform limitations (e.g., communication channels, predefined incentives and metrics);
  • 8. 8 CoopIS’13 3) Experimental evaluation (e.g., on micro-task platforms, with students, volunteers) Existing Evaluation Approaches PRO CON controlled environment and reproducible setups platform limitations (e.g., communication channels, predefined incentives and metrics); limited monetary funds may derive skewed results; working with people inherently willing or forced to perform work may derive skewed results
  • 9. 9 CoopIS’13  Offer methodology for quickly selecting, composing and customizing existing incentive mechanisms.  Roughly predicting effects of composition in dynamic crowdsourcing environments.  Model and simulation parameters can be changed dynamically, allowing quick testing of different incentive scheme setups and behavioral responses at low cost.  Modeling of incentives and responses of arbitrary complexity.  We do not devise novel nor optimal incentive mechanisms! Our approach
  • 10. 10 CoopIS’13 Simulation Model of Incentive Mechanism Decision-making function fa considers: 1) the statistically or intentionally determined personality of the worker St 2) historical records of past actions {S0, … , St-1} 3) authority’s view of worker’s performance Mj 4) performance of other workers {Mk}, k ≠ j 5) promised rewards R Incentive mechanism IM considers: 1) current state of artifact Ki 2) the current performance metrics of worker Mj 3) output from another incentive mechanism returning the same type of reward R′ak
  • 11. 11 CoopIS’13  True power of incentives  composition of incentive mechanisms  Two basic operators on incentive mechanisms: – addition (+) and functional composition () – operate on common metrics – final metrics’ values advertised to workers represent the promised reward  Major difficulty in designing successful incentive strategies is to properly choose performance metrics, basic incentive mechanisms and the proper composition. Simulating Complex Incentive Strategies
  • 12. 12 CoopIS’13 Simulation Methodology 1) Defining domain-specific meta-model by extending core meta-model 2) Capturing worker behavioral patterns and reward calculation into executable model 3) Defining scenarios, assumptions, and configurations for individual simulation runs 4) Evaluating and interpreting simulation results 1 2 3 4
  • 13. 13 CoopIS’13 Simulation Methodology 1) Defining domain-specific meta-model by extending core meta-model 2) Capturing worker behavioral patterns and reward calculation into executable model 3) Defining scenarios, assumptions, and configurations for individual simulation runs 4) Evaluating and interpreting simulation results 1 2 DomainPro Designer *www.quandarypeak.com
  • 14. 14 CoopIS’13 Simulation Methodology 1) Defining domain-specific meta-model by extending core meta-model 2) Capturing worker behavioral patterns and reward calculation into executable model 3) Defining scenarios, assumptions, and configurations for individual simulation runs 4) Evaluating and interpreting simulation results 3 DomainPro Analyst 4 *www.quandarypeak.com
  • 15. 15 CoopIS’13  Simulation model of a realistic scenario, inspired by – Citizen-driven traffic reporting (SmartJourney – Aberdeen) – Crowdsourced software testing  Generalized scenario: – Entities: Authority, Workers, Situations, Reports – Activities: Submit, Improve, Rate, Report duplicates – Metrics: Reputation (for trustworthiness), Points (for productivity) – Incentives: Three incentive mechanisms:  IM1 – fixed amounts of points per activity  IM2 – points related with report quality  IM3 – users are assigned reputation based on past activities Evaluation
  • 16. 16 CoopIS’13  Composite Incentive Schemes (CIS) evaluated:  3 Experiments: – Exp1: Compare impact of CIS1, CIS2, and CIS3 on authority cost. – Exp2: Analyze effects of having too few or too many workers per situation – Exp3: Evaluate effects of malicious workers (0-50%) on cost. Evaluation
  • 17. 17 CoopIS’13  CIS3 most reasonable to use. Can cope well with up to 20% of malicious workers. Evaluation Results – Example
  • 18. 18 CoopIS’13  Presented a methodology for modeling and simulating incentives in crowdsourcing environments.  Advantages: – Useful for quick, runtime, approximate evaluations of different compositions of incentive mechanisms.  Drawbacks: – Inconclusive results. See: Advantages  Future Work: – Devise suitable models for more complex socio-technical systems. Conclusion & Outlook
  • 19. 21st International Conference on Cooperative Information Systems (CoopIS’13) September 11-13, 2013, Graz, Austria Ognjen Scekic, Christoph Dorn, Schahram Dustdar Distributed Systems Group Vienna University of Technology http://dsg.tuwien.ac.at Thank you! Questions? 21st International Conference on Cooperative Information Systems (CoopIS’13) September 11-13, 2013, Graz, Austria