A Clustering-Based Approach for Discovering Interesting Places in a Single Trajectory

With the development of many location sensors such as GPS technology and mobile communication devices, a lot of trajectories of users and moving objects can be obtained. These trajectories may contain many interesting individual patterns of the users and moving objects. This creates an appropriate basis for developing efficient new methods for mining moving objects. Semantic clustering of trajectories left behind moving objects is an important aspect in spatio-temporal data mining. Algorithm CB-SMoT (Clustering-Based Stops and Moves of Trajectories) is based on a traditional algorithm DBSCAN which is a classical density-based clustering approach. There is an important parameter Eps in the algorithm CB-SMoT, whose value can dramatically affect the quality of clustering. With in-depth analysis of spatial and temporal characteristics of trajectory data and some related statistical theory, the trajectory data is pre-processed. A new method of calculating the Eps value is proposed. The experiment proves that using this method to calculate the parameter values can significantly improve the quality of clustering.