21 November 2016 - Associate Professor Roger Zimmermann from the Department of Computer Science and two of his students, Yin Yifang and Rajiv Ratn Shah, have won the Best Paper Award for their paper titled "A General Feature-Based Map Matching Framework with Trajectory Simplification".

The paper was presented at the 7th ACM SIGSPATIAL International Workshop on GeoStreaming, which was held in conjunction with the 24th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems in San Francisco.

Speaking on behalf of the team, Yin Yifang said, “Traditional map matching techniques are effective when dealing with trajectories where the GPS sampling interval is less than 30 seconds. However, the sampling interval of real GPS data varies significantly from less than one minute to more than five minutes in real world applications. Also, due to practical considerations, the current GPS sampling interval is usually one minute or more, given that a shorter interval will entail a larger amount of data that needs to be saved.” The low sampling rate results in lower accuracy in predicting the route that a driver takes. The team examined how the uncertainty in the trajectory taken by a vehicle can be reduced by matching the GPS points to the road network on a digital map. Aside from GPS data, the paper takes factors such as road length and transition angle into consideration in increasing the accuracy of predicting the path taken by a vehicle.

Yifang added that it is commonly assumed that a driver always chooses to take the shortest path between two points, but she said this is not always true. “Given two routes, drivers may still choose to take the longer route due to safety concerns if the alternative involves more turns. The challenge we had to face in our research was how to factor in all the considerations that impact a driver's choice of route.” She said that in future, factors like speed limit and the presence of road bumps along a route can be integrated into the GPS to assist in route prediction.