Search

Task assignment is the fundamental problem of crowdsourcing, which addresses the matching between service providers and demanders in urban services. As crowdsourcing has been receiving more attention, the focus of task assignment has expanded from one-sided or two-sided task assignment to three-sided task assignment. Recent research on the three-sided version aims at the online case or stable case, which only partially reflects the real scenario. In this paper, we consider a hybrid case that contains the aforementioned cases. Concretely, we propose a three-sided online stable task assignment problem. We first show the fundamental notations of task assignment. Then, we analyze and present the matching conditions and formulate the problem. As a warm-up, we propose three iconic algorithms as baselines and analyze their performance in the three-sided scenario. Inspired by two suggested matching technique in a two-sided task assignment problem, we present a new algorithm that divides the arrival sequence into three parts in which we utilize different strategies for different arrival intervals. It balances efficiency and effectiveness. In the experiment, all the algorithms are verified in size, memory, and time and tested on a real and a synthetic dataset. The results show that our algorithm maintains efficiency and effectiveness. [ABSTRACT FROM AUTHOR]