Towards a Better Understanding of Context and Context-Awareness
HUC '99 Proceedings of the 1st international symposium on Handheld and Ubiquitous Computing
Using GPS to learn significant locations and predict movement across multiple users
Personal and Ubiquitous Computing
Design of a Car Navigation System that Predicts User Destination
MDM '06 Proceedings of the 7th International Conference on Mobile Data Management
Learning and inferring transportation routines
Artificial Intelligence
Traffic-Known Urban Vehicular Route Prediction Based on Partial Mobility Patterns
ICPADS '09 Proceedings of the 2009 15th International Conference on Parallel and Distributed Systems
A system for destination and future route prediction based on trajectory mining
Pervasive and Mobile Computing
Vehicular networks and the future of the mobile internet
Computer Networks: The International Journal of Computer and Telecommunications Networking
Evaluating Application Prototypes in the Automobile
IEEE Pervasive Computing
Hi-index | 0.00 |
A route trajectory is described as the upcoming course of the road in geographical terms as well as in terms of time. This enables the possibility of providing context-aware applications in modern vehicles. In order to generate such a trajectory with its context aware information, it is still necessary to enter a destination point into a navigation system. However, the most frequent commutations (drives, trips) are to known destinations and are therefore not performed with any active guidance system. This means that the prediction must be determined in a different manner. This paper presents a method, which predicts the route trajectory of a vehicle based on the travel history of its user. In addition to the traveled distance further context parameters are used for the prediction. These parameters include the current time of day, day of week and the route frequency, which indicates the number of times a particular route has already been traveled. Moreover, the developed prediction is evaluated in a volunteers study with about 500 rides and about 9.500 driven kilometers. The results show that in 80 percent of the cases the forward-lying path can be predicted correctly.