On Predicting the Future Locations of Moving Objects: The State of the Art