Exploring Diversified Incentive Strategies for Long-Term Participatory Sensing Data Collections

The popularity of smartphones provides a new data collection paradigm, participatory sensing, which extends existing web-based crowdsourcing applications to pervasively collect sensing data around the physical world. It is important to design an effective incentive strategy to attract enough users' participation to guarantee the success of participatory sensing applications. Most of existing incentive mechanisms lack sufficient evaluation on the feasibility and effectiveness in a practical application scenario. Although some recent studies consider several practical factors in psychology and sociology, and evaluate different incentive strategies in practical experiments, but still fail to consider the time and location dependence of sensing data or the long-term sensing data collection requirements. In this paper, we focus on a new participatory sensing application, namely to recruit users to take time and geo-tagged images of the sky for a long duration to build a large-scale dataset, which is used in the research on outdoor air quality level inference. Three incentive strategies, linear reward, competitive and random red envelope are considered. Furthermore, we explore to combine three incentive strategies in different orders to investigate their effects on long-term data collections and perform practical experiments in three universities for a long duration (6 weeks). A number of sky images are collected, from which we obtain some interesting insights, providing important guidelines for designing diversified incentive strategies for long-term participatory sensing data collections.