Survey of Trajectory Mining using Uncertain Sensor Data Model with Probabilistic Suffix Tree

Abstract

This paper describes a framework that works by collecting the trajectory data obtained from the sensors. The data is stored and processed in a way that helps in identifying events such as key activity areas, evolving activity, etc. It helps to attain better insight into the work habits of the population. Trajectory mining is either assumed that the timeordered location data recorded as trajectories are either deterministic or that the uncertainty, e.g., due to equipment or technological limitations, is removed by incorporating some pre-processing routines. Thus, the trajectories are processed as deterministic paths of mobile object location data. Probabilistic trajectory extraction and mining from uncertain trajectory data is the first phase analysis on the subject. It is also interested in identifying and developing alternative approaches with the use of which can make the approach more scalable, e.g. a trajectory compression scheme could be developed to further decrease the length of the trajectories. This paper proposes an efficient distributed mining algorithm to jointly identify a group of sensor data and discover their trajectory of sensor data in wireless sensor networks. Then, Map-Reduce algorithm (Probabilistic Suffix Tree) is introduced which utilizes the discovered group trajectory sensor data shared by the transmitting node.