Spatial Task Allocation Based on User Trajectory Prediction

Abstract

With the rapid development of wireless communication technology and mobile intelligent terminals, location based service has been widely used for its unique features such as mobility, practicality and portability, including mobile crowdsourcing. In mobile computing environment, mobile crowdsourcing task allocation has become a focused research issue. In mobile crowdsourcing, application scenarios are in dynamic state, and workers are also willing to accept tasks. In response to these challenges, this paper proposes a mobile crowdsourcing task allocation strategy based on user trajectory prediction. First, the location points in user historical trajectory data are clustered into regions using k-means algorithm. Then the user’s trajectory is analyzed and excavated to get the user’s mobile pattern. On this basis, we extract mobile rules and calculate the confidence. According to the mobile rules, we predict the region that the user will reach, and finally assign the tasks in the region to him. The prediction based task allocation method proposed in this paper avoids the additional cost that platform need to pay to users, recommends user the task that more suitable for him, and improves the success rate of task allocation. Finally, based on the analysis and simulation experiments of real datasets, the proposed method can effectively predict the location region of users, and at the same time, it can achieve better results than other methods in the situation of task allocation and completion.