Task-Scheduling Based on Multi-Objective Particle Swarm Optimization in Spatial Crowdsourcing

As a result of the rapid growth of internet and smartphone technology, a novel platform that attracts individuals and groups known as crowdsourcing emerged. Crowdsourcing is an outsourcing platform that facilitates the accomplishment of costly tasks that consume long periods of time when traditional methods are used. Spatial crowdsourcing (SC) is based on location; it introduces a new framework for the physical world that enables a crowd to complete spatialtemporal tasks. The primary issue in SC is the assignment and scheduling of a set of available tasks to a set of proper workers based on different factors, such as the location of the task, the distance between task location and hired worker location, temporal conditions, and incentive rewards. In the real-world, SC applications need to optimize multi-objectives simultaneously to exploit the utility of SC, and these objectives can be in conflict. However, there are few studies that address this multi-objective optimization problem within a SC environment. Thus, the authors propose a multi-objective task scheduling optimization problem in SC that aims to maximize the number of completed tasks, minimize total travel cost, and ensure worker workload balance. To solve this problem, we developed a method that adapts the multi-objective particle swarm optimization (MOPSO) algorithm based on a proposed novel fitness function. The experiments were conducted with both synthetic and real datasets; the experimental results show that this approach provides acceptable initial results. As future work, we plan to improve the effectiveness of our proposed algorithm by integrating a simple ranking strategy based on task entropy and expected travel costs to enhance MOPSO performance.