Key points are not available for this paper at this time.
Mobile Crowd Sourcing (MCS) has been an enabler in the development of artificial intelligence (AI) in general, and machine learning in particular. From collecting data to giving meaning to the data, there has been considerable work supporting the use of MCS in AI. While successful, current MCS solutions still suffer from limitations such as workers recruitment, data quality, trust, and so on, that can benefit a great deal from AI. However, the integration of AI in MCS is still at a nascent stage, thus opening various opportunities for further research. In this article, we review and discuss the integration of AI in MCS solutions, highlight its research challenges, and suggest means to address them. We also propose a novel architecture for AI-based MCS, where AI techniques are integrated and embedded in the different layers of MCS framework to provide efficient and trusted MCS applications. In particular, a machine learning (ML)-based selection using behaviors of individual workers is proposed, and its efficacy is gauged by analyzing a use-case study. The results show that by implementing a hybrid approach, the efficiency of selection was considerably improved. This article demonstrates a clear overview of AI-based MCS solutions and provide guidelines on applying AI to solve the current challenges and open issues.
Abououf et al. (Tue,) studied this question.