Knowledge Discovery on Trajectory Data Warehouses: Possible usage of the Data Mining Techniques
Fernando José Braz · 2008
In this paper we are interested in discussing the possibility of theusage of Data Mining tasks in order to reveal knowledge resident in Trajectory Data Warehouses (TDW). We consider a data stream environment where a set of mobile objects send the data about its location in a irregular and unbounded way. The data volume is stored in a centralized and traditional DW with precomputed aggregations values (preserving the trajectories privacy). Through of analysis of the TDW measures (pre-computed aggregation values) we can reveal some characteristics about trajectories in a given spatio-temporal area. The revealed knowledge can be useful in order to describe or show the occurrence of a real phenomenon. We present a review of a proposed structure of a TDW and discuss the use of Data Mining tasks to improve the analysis of the trajectory data warehouse environment.