Collusion-resistant Worker Selection in Social Crowdsensing Systems

Abstract

The main idea behind social crowdsensing is to leverage social friends as crowdworkers to participate in crowdsensing tasks. A main challenge, however, is the identification and recruitment of well-suited workers. This becomes especially more challenging for large-scale online social networks with potential sparseness of the friendship network which may result in recruiting participants who are not in direct friendship relations with the requester. Such recruitment may increase the possibility of collusion among participants, thus threatening the application security and affecting data quality. In this paper, we propose a collusion-resistant worker selection method which aims to prevent the selection of colluders as suitable participants. For each participant who is considered to be selected as suitable, the proposed method is aimed to prevent any possible collusion. To do so, it determines whether the selection of a new participant may result in the formation of a colluding group among the selected participants. This has been achieved through leveraging the Frequent Itemset Mining technique and defining a set of collusion behavioral indicators. Simulation results demonstrate the efficacy of our proposed collusion prevention method in terms of selecting efficient collusion indicators and detecting the colluding groups.