Adaptive task scheduling in service oriented crowd using SLURM

Abstract

Crowdsourcing is a distributed problem-solving paradigm. Service oriented crowdsourcing paradigm involves both consumers and service providers. A consumer requests for a service (task); a provider provides that service (does that task); and the providers are paid by consumers for the service as per their satisfaction. The challenge is to select a service provider from a list of providers which can provide maximum satisfaction to the consumer for that service. This work outlines an architectural model using SLURM tool for efficient management of crowd. At the center of this work, we proposed a novel idea of adaptive task scheduling which is based on the customer satisfaction feedbacks. Our approach improves efficiency, and decreases the cost of service to consumers. Experimental results demonstrate the viability of our approach. © 2013 Springer-Verlag Berlin Heidelberg.