Representing positional uncertainty of individual and aggregated trajectories of moving objects

Trajectories are used to represent objects' movement and spatio-temporal aggregation of trajectories is commonly used in knowledge discovery. Based on a general workflow to extract knowledge, we identify relevant factors that propagate uncertainty in moving object datasets when using aggregation. We use a probabilistic approach and propose a theoretical model to represent positional uncertainty of trajectories and their aggregation, and implement a prototype system to compute positional uncertainty in pedestrians' movement data.