10th IEEE International Conference on Collaborative Computing: Networking, Applications and Worksharing

Research Article

Combining Human and Machine Computing Elements for Analysis via Crowdsourcing

Download227 downloads
  • @INPROCEEDINGS{10.4108/icst.collaboratecom.2014.257298,
        author={Brian Blake and Julian Jarrett and Iman Saleh and Rohan Malcolm and Sean Thorpe and Tyrone Grandison},
        title={Combining Human and Machine Computing Elements for Analysis via Crowdsourcing},
        proceedings={10th IEEE International Conference on Collaborative Computing: Networking, Applications and Worksharing},
        publisher={IEEE},
        proceedings_a={COLLABORATECOM},
        year={2014},
        month={11},
        keywords={crowdsouring; experimentation; elastic systems},
        doi={10.4108/icst.collaboratecom.2014.257298}
    }
    
  • Brian Blake
    Julian Jarrett
    Iman Saleh
    Rohan Malcolm
    Sean Thorpe
    Tyrone Grandison
    Year: 2014
    Combining Human and Machine Computing Elements for Analysis via Crowdsourcing
    COLLABORATECOM
    IEEE
    DOI: 10.4108/icst.collaboratecom.2014.257298
Brian Blake1, Julian Jarrett1,*, Iman Saleh1, Rohan Malcolm2, Sean Thorpe2, Tyrone Grandison3
  • 1: University of Miami
  • 2: University of Technology, Jamaica
  • 3: Proficiency Labs
*Contact email: j.jarrett@miami.edu

Abstract

Crowd computing leverages human input in order to execute tasks that are computationally expensive, due to complexity and/or scale. Combined with automation, crowd computing can help solve problems efficiently and effectively. In this work, we introduce an elasticity framework that adaptively optimizes the use of human and automated software resources in order to maximize overall performance. This framework includes a quantitative model that supports elasticity when performing complex tasks. Our model defines a task complexity index and an elasticity index that is used to aid in decision support for assigning tasks to respective computing elements. Experiments demonstrate that the framework can effectively optimize the use of human and machine computing elements simultaneously. Also, as a consequence, overall performance is significantly enhanced.