Results 1 -
5 of
5
Faster Batched Shortest Paths in Road Networks
- ATMOS
, 2011
"... We study the problem of computing batched shortest paths in road networks efficiently. Our focus is on computing paths from a single source to multiple targets (one-to-many queries). We perform a comprehensive experimental comparison of several approaches, including new ones. We conclude that a new ..."
Abstract
-
Cited by 9 (5 self)
- Add to MetaCart
We study the problem of computing batched shortest paths in road networks efficiently. Our focus is on computing paths from a single source to multiple targets (one-to-many queries). We perform a comprehensive experimental comparison of several approaches, including new ones. We conclude that a new extension of PHAST (a recent one-to-all algorithm), called RPHAST, has the best performance in most cases, often by orders of magnitude. When used to compute distance tables (many-to-many queries), RPHAST often outperforms all previous approaches.
Efficient Frequent Sequence Mining on Taxi Trip Records Using Road Network Shortcuts
"... Huge amounts of geo-referenced spatial location data and moving object trajectory data are being generated at ever increasing rates. Patterns discovered from these data are valuable in understanding human mobility and facilitating traffic mitigation. In this study, we propose a new approach to minin ..."
Abstract
-
Cited by 1 (1 self)
- Add to MetaCart
Huge amounts of geo-referenced spatial location data and moving object trajectory data are being generated at ever increasing rates. Patterns discovered from these data are valuable in understanding human mobility and facilitating traffic mitigation. In this study, we propose a new approach to mining frequent patterns from large-scale GPS trajectory data after mapping GPS traces to road network segments. Different from applying association rule-based frequent sequence mining algorithms directly, which generally have high computation overhead and are not scalable, our approach utilizes hierarchies of road networks. After contracting nodes and creating shortcuts by contraction hierarchies algorithms, the original road segment sequences are transformed into sequences of shortcuts with much smaller data volumes. By using computed shortest paths as simulated GPS trajectories, our experiments on 17,558 selected taxi trip records in NYC in January 2009 have shown that runtimes of frequent sequence mining on shortcut sequences are orders of magnitude faster than on original road segment sequences. In addition, frequent subsequences in shortcuts are more informative and interpretable based on the betweenness centralities of the shortcuts than visualizing betweenness centralities of individual road segments. 1
Compass-Based Navigation in Street Networks
"... Abstract. We present a new method for navigating in a street network using solely data acquired by a (smartphone integrated electronic) compass for self-localization. To make compass-based navigation in street networks practical, it is crucial to deal with all kinds of imprecision and different dri ..."
Abstract
- Add to MetaCart
(Show Context)
Abstract. We present a new method for navigating in a street network using solely data acquired by a (smartphone integrated electronic) compass for self-localization. To make compass-based navigation in street networks practical, it is crucial to deal with all kinds of imprecision and different driving behaviors. We therefore develop a trajectory representation based on so-called inflection points which turns out to be very robust against measurement variability. To enable real-time localization with compass data, we construct a custom-tailored data structure inspired by algorithms for efficient pattern search in large texts. Our experiments reveal that on average already very short sequences of inflection points are unique in a large street network, proving that this representation allows for accurate localization.
Proceedings of the Twenty-Third International Joint Conference on Artificial Intelligence Map Matching with Inverse Reinforcement Learning
"... We study map-matching, the problem of estimating the route that is traveled by a vehicle, where the points observed with the Global Positioning System are available. A state-of-the-art approach for this problem is a Hidden Markov Model (HMM). We propose a particular transition probability between la ..."
Abstract
- Add to MetaCart
We study map-matching, the problem of estimating the route that is traveled by a vehicle, where the points observed with the Global Positioning System are available. A state-of-the-art approach for this problem is a Hidden Markov Model (HMM). We propose a particular transition probability between latent road segments by the use of the number of turns in addition to the travel distance between the latent road segments. We use inverse reinforcement learning to estimate the importance of the number of turns relative to the travel distance. This estimated importance is incorporated in the transition probability of the HMM. We show, through numerical experiments, that the error of map-matching can be reduced substantially with the proposed transition probability. 1
and
"... This paper proposes a kind of location-based research, human travel route prediction, which is to predict the track of a subject’s future movements. The proposed method works as follows. The mobile user sends his/her current route along with several dummy routes to the server by using a 3D route mat ..."
Abstract
- Add to MetaCart
(Show Context)
This paper proposes a kind of location-based research, human travel route prediction, which is to predict the track of a subject’s future movements. The proposed method works as follows. The mobile user sends his/her current route along with several dummy routes to the server by using a 3D route matrix, which encodes a set of routes. The server restores the routes from the 3D matrix and matches the restored routes to the saved routes. The predicted route is found as the trunk of the tree, which is built by superimposing the matching results. The server then sends the predicted routes back to the user, who will apply the predicted route to a real-world problem such as traffic control and planning. Preliminary experimental results show the proposed method successfully predicts human travel routes based on current and previous routes. User privacy is also rigorously protected by using a simple method of dummy routes.
