How Mobile Contributors Will Interact With Each Other in Mobile Crowdsourcing With Word of Mouth Mode

Abstract

Mobile crowdsourcing is a promising paradigm for collecting sensing data by leveraging contributions of numerous mobile smart phones. It works efficiently with Word of Mouth Mode (WoM), especially for sensing tasks with time and location constraints, since the sensing task can be spread quickly among mobile contributors in the WoM mode. In this paper, we first investigate the behaviors of contributors, categorize all contributors into four types according to their different behaviors, and propose an inviting algorithm for contributors to recruit cooperators through social closeness. Then, we design a reward mechanism for crowdsourcing platform to evaluate the budget and pay the reward to contributors, meanwhile stimulate contributors to make the maximum contribution. Furthermore, considering two different scenarios, we model the interactions among contributors as two Stackelberg games, and backward induction approach is used to analyze each game. We propose an algorithm to compute the best response of every contributor, and we theoretically prove that this proposed algorithm may converge a unique Stackelberg equilibrium. The proposed approach can be applied to task formulation and task budget evaluations for crowdsourcing platforms.