Abstract
The ubiquity of mobile platforms and smartphones breeds a large number of spatial crowdsourcing applications like TaskRabbit and Uber. In spatial crowdsourcing, workers are financially motivated to perform as many self-selected tasks as possible to maximize their revenue. In spatial crowdsourcing, obtaining an optimal or near-optimal task schedule for a worker to accomplish as many tasks as possible is a crucial yet quite challenging problem. In this project, we present our solution to task assignments among multiple workers in spatial crowdsourcing using Particle Swarm Optimization (PSO), which is based on the concept of “Flocking of Birds”. The goal of this project is to propose a new task assignment framework based on PSO with the objective of maximizing the number of spatial tasks that can be completed by a given number of workers in spatial crowdsourcing. Extensive experiments using the New York City’s Foursquare check-in dataset showed that our proposed PSO-based framework outperforms the baseline approach for both short and long tasks in terms of the number of tasks that can be completed by a worker group. The parameter sensitivity analysis of our proposed PSO-based framework was also provided.