Crowdsourcing defines the process of outsourcing jobs to an unknown group of people. In recent years many platforms evolved which realize this concept. These platforms act as mediator between a company or person (the employer), who creates the jobs or tasks and the crowd (the workers) which work on the tasks and fulfill them for a reward. Until today, crowdsourdcing platforms mediate thousands of tasks to thousands of workers, resulting in a growing economical and social importance.
Crowdsourcing platforms vary concerning the offered workflows and processes as well as the supported types of tasks. The task types are ranging from difficult research tasks and creative tasks to so called micro tasks. Micro tasks are often simple, repetitive, e.g. data annotation, translation or surveys, and require only a short completion time. However, even if micro-jobs are rather simple from human beings they are yet still impossible to complete by algorithmic approaches.
This project focuses on designing and evaluating new mechanisms in micro tasking platforms to improve the basic concepts with respect to the interests of provider, employer and worker. The mechanisms should simplify and speed up the organizational processes and increase the quality of the working results. Thus, deepening the knowledge and understanding about the workflow, the task design and the interests of all parties, developing quality management processes and optimizing the task worker mapping by the integration of recommendation systems are parts of the project. We use various methods and metrics of different scientific fields, e.g. data mining, modeling theory and psychology.